🧊 3D Vision

🎞️ ECCV2026 · 512 paper notes

📌 Same area in other venues: 📷 CVPR2026 (751) · 🔬 ICLR2026 (195) · 🧪 ICML2026 (30) · 🤖 AAAI2026 (79) · 🧠 NeurIPS2025 (116) · 📹 ICCV2025 (267)

🔥 Top topics: 3D Gaussian Splatting ×97 · 3D Reconstruction ×40 · Dynamic Scenes ×32 · Novel View Synthesis ×26 · Diffusion Models ×23

RoMa v2: Harder Better Faster Denser Feature Matching

RoMa v2 combines multi-view coarse matching over frozen DINOv3 features, separately trained efficient refiners, and pixel-wise uncertainty prediction to improve challenging-viewpoint matching and subpixel precision, reducing RoMa's AerialMegaDepth EPE from 25.05 to 4.12 pixels without claiming universal superiority in extreme cross-modal generalization or absolute speed.

Ultra3D: Efficient and High-Fidelity 3D Generation with Part Attention

Ultra3D first establishes a coarse object layout with compact VecSet tokens, then generates sparse voxel latents using attention organized along part boundaries, supporting 1024-resolution meshes with 6.7× faster part self-attention and 3.3× faster DiT inference than a full-attention baseline.

P3-SAM: Native 3D Part Segmentation

P3-SAM learns a single-point-prompted, two-stage multi-mask network from approximately 3.7 million native 3D assets and combines quality prediction with deduplication for automatic segmentation, reaching an average IoU of 59.88 on PartObj-Tiny without connectivity, compared with PartField's 53.93.

VolSplat: Rethinking Feed-Forward 3D Gaussian Splatting with Voxel-Aligned Prediction

VolSplat shifts the primitive prediction paradigm of feed-forward 3D Gaussian Splatting from pixel-aligned to voxel-aligned: it first projects multi-view 2D features back to 3D space to form a voxel feature grid through depth maps, refines them using residual learning via a sparse 3D U-Net, and then predicts Gaussian parameters per voxel, thus fundamentally decoupling the 3D representation from the rigid binding of the input image resolution to achieve adaptive Gaussian density allocation.

LoGeR: Long-Context Geometric Reconstruction with Hybrid Memory

LoGeR reconstructs video chunk by chunk, preserving adjacent-chunk detail through sliding-window attention and propagating global geometry through test-time-updated fast weights; its feedforward-aligned LoGeR* variant reduces average KITTI ATE from TTT3R's 72.86 m to 18.65 m, although very-long-sequence evaluation includes periodic state resets.

MVGS: Multi-view Regulated Gaussian Splatting for Novel View Synthesis

MVGS replaces single-view iterative fitting in Gaussian Splatting with joint multi-view updates, cross-intrinsic guidance, and two densification strategies, improving 3DGS PSNR on Mip-NeRF 360 from 28.69 to 29.61 while increasing training time from 0.3 to 2.4 hours.

Skyfall-GS: Synthesizing Immersive 3D Urban Scenes from Satellite Imagery

Skyfall-GS anchors coarse urban geometry and appearance in satellite observations, then progressively transfers diffusion-generated facade details into a 3D Gaussian scene through a descending view curriculum, achieving FIDCLIP 27.03 and CMMD 2.110 on DFC2019 and enabling real-time exploration after offline optimization.

XYZ-IBD: Benchmarking Robust 6D Object Pose Estimation under Real-World Industrial Complexity

XYZ-IBD combines multi-sensor real-world bin data, coating-assisted geometric annotation, and noise-matched simulation to benchmark industrial 6D pose estimation across 15 parts, exposing a substantial performance gap: SAM6D achieves 0.578 6D AP on XYZ-IBD versus its BOP-Core 5 average of 0.704.

Generation Models Know Space: Unleashing Implicit 3D Priors for Scene Understanding

VEGA-3D repurposes frozen Wan2.1 as a spatial feature extractor, combining moderate noise injection, intermediate-layer features, and token-level gated fusion to raise ScanRefer [email protected] from 51.7 to 56.2, with varying gains on spatial reasoning and manipulation.

LoMa: Local Feature Matching Revisited

LoMa retains the basic division between the DaD detector, DeDoDe descriptor, and LightGlue matcher, but scales training data, updates the training recipe, and widens the matcher to reach 54.3 mAA@10px on the new HardMatch test set, 18.6 points above ALIKED+LightGlue.

Register Any Point: Scaling 3D Point Cloud Registration by Flow Matching

RAP reformulates multi-view point cloud registration as point-coordinate generation conditioned on local geometry, stabilizes flow sampling with stepwise rigidity enforcement, and recovers per-view poses from generated points, achieving 85.9% edge success, 77.3% graph success, and an average runtime of 8.9 seconds on the authors' zero-shot cross-domain benchmark.

RayRoPE: Projective Ray Positional Encoding for Multi-view Attention

RayRoPE represents image patches as ray segments in the query camera frame and computes expected RoPE over uncertain predicted depths, improving multi-view geometry without additional supervision of internal depths and reducing small-scale LVSM's RE10K LPIPS from 0.112 with PRoPE to 0.085.

AnyView: Synthesizing Any Novel View in Dynamic Scenes

AnyView feeds a monocular video and source/target camera rays into a pretrained video diffusion model, learning implicit spatiotemporal relationships from cross-domain multi-view data without depth reprojection or test-time optimization, and achieves 12.13 dB average PSNR and 1757 FVD on zero-shot extreme-view AnyViewBench tests without guaranteeing that hallucinated occluded content is correct.

DINO-SLAM: DINO-Informed RGB-D SLAM for Neural Implicit and Explicit Representations

DINO-SLAM uses an online-optimized scene geometry encoder to fuse frozen DINOv2 features with depth into geoDINO, then integrates these features into implicit or explicit maps, raising its GauS-SLAM-based variant's average PSNR across eight Replica scenes from 40.25 to 41.42 dB while improving depth and trajectory errors.

ReSplat: Learning Recurrent Gaussian Splatting

ReSplat predicts a compact Gaussian scene from posed images and recurrently corrects its Gaussians and hidden states using input-view rendering errors, improving eight-view DL3DV reconstruction from 26.21 dB at initialization to 27.70 dB without explicit test-time gradient computation while retaining approximately 1/16 of the Gaussian count of per-pixel baselines.

UniPR-3D: Towards Universal Visual Place Recognition with Visual Geometry Grounded Transformer

UniPR-3D introduces the 3D geometry-aware tokens of VGGT into visual place recognition for the first time. It designs tailored aggregation strategies for the distinct characteristics of 2D and 3D tokens and supports both single-frame and variable-length sequence matching, outperforming existing single-view and multi-view methods on multiple benchmarks.

VGGT-World: Transforming VGGT into an Autoregressive Geometry World Model

VGGT-World treats intermediate frozen VGGT features as world states, learns their future evolution with clean-target prediction and a two-stage flow-forcing curriculum, and jointly decodes them with the original geometry heads, reaching 0.065 short-term KITTI AbsRel without generating future RGB frames.

Flow4R: Unifying 4D Reconstruction and Tracking with Scene Flow

Flow4R uses a shared two-view Transformer to predict local geometry and cross-frame point positions, then separates camera and object motion through a learned-weight geometric solver, unifying 4D reconstruction and tracking without global post-optimization and achieving 71.1 all-point tracking APD and 81.00 reconstruction APD on Point Odyssey.

GlobalSplat: Efficient Feed-Forward 3D Gaussian Splatting via Global Scene Tokens

GlobalSplat aligns multiple views into a fixed set of global scene tokens before decoding compact 3D Gaussians, achieving 28.53 dB PSNR, 77.88 ms per forward pass, and a 3.8 MB scene asset with 16K Gaussians on 24-view RealEstate10K evaluation, while retaining less perceptual detail than the heavier ZPressor.

InverseCrafter: Efficient Video ReCapture as a Latent Domain Inverse Problem

InverseCrafter formulates monocular video recapture along a specified camera trajectory as latent-space inpainting, predicts continuous multichannel masks with a lightweight encoder, and enforces measurement consistency through back-propagation-free DDS, achieving a runtime of 71 seconds and measurement PSNR of 29.35 dB without fine-tuning the video diffusion model.

MetricAnything: Scaling Metric Depth Pretraining with Noisy Heterogeneous Sources

MetricAnything organizes pretraining on roughly 20 million heterogeneous image-depth pairs through random sparse depth prompts, then distills dense labels from a conditioned teacher into RGB-only students, achieving strong average rankings in depth completion and monocular metric depth while showing scaling gains that depend jointly on data, prompting, and distillation design.

Repurposing Geometric Foundation Models for Multi-view Diffusion

GLD repurposes frozen Depth Anything 3 features as a multi-view diffusion space, combining joint boundary-level generation, frozen deeper-feature propagation, and a shallow cascade to achieve Re10K PSNR 16.362 without text-to-image generative pretraining while directly reusing geometry decoding for depth and ray maps.

SLAM-Former: Putting SLAM into One Transformer

SLAM-Former alternates causal incremental mapping and global map refinement with a shared-weight Transformer, feeding refined KV caches back to the frontend to achieve a mean TUM RGB-D ATE of 0.039 m and a 7-Scenes reconstruction Chamfer distance of 0.027 m without camera calibration.

Track4World: Feedforward World-centric Dense 3D Tracking of All Pixels

Track4World uses global video geometry to support 2D-to-3D correlation decoding and builds world-centric trajectories for every pixel of every frame from queryable pairwise scene flow, achieving an EPE3D of 0.0742 on KITTI without extending expensive 3D neighborhood searches directly to all pixels.

Tri-Efficient Transfer Learning for Point Cloud Videos

PoinTriE learns motion priors through rigid pseudo motion and cross-modal alignment of static point clouds, then combines a frozen backbone, a low-rank side network, and unit masking to achieve 94.37% average accuracy across four clip-length settings on MSR-Action3D while tuning only 2.2% of parameters as reported in Table 1.

VIGS-SLAM: Visual Inertial Gaussian Splatting SLAM

VIGS-SLAM jointly optimizes dense visual correspondences and IMU constraints, stabilizes startup through staged initialization, and propagates loop corrections to its Gaussian map, achieving 2.79 cm average ATE RMSE on EuRoC while improving novel-view rendering in difficult scenes.

One4D: Unified 4D Generation and Reconstruction via Decoupled LoRA Control

One4D encodes varying numbers of observed frames through unified masked conditioning and jointly generates appearance and geometry using separately computed RGB/XYZ LoRA branches with shared frozen video weights and sparse bidirectional control links, supporting single-image 4D generation, sparse-frame completion, and full-video reconstruction in one model.

Reconstruction by Generation: 3D Multi-Object Scene Reconstruction from Sparse Observations

RecGen jointly generates complete shapes and camera-frame poses from one or two RGB-D views and instance masks, then conditions texture generation on pose, raising strict pose accuracy on ArtVIP parts from SAM3D's 45.8% to 84.0%, although two-view reconstruction and appearance metrics do not uniformly outperform the alternatives.

SplatPainter: Interactive Authoring of 3D Gaussians from 2D Edits via Test-Time Training

SplatPainter compresses an existing 3D Gaussian asset into contextualized voxel latents, then uses test-time training to absorb 2D edits and update spatially selected parts of the asset, substantially reducing appearance drift in unedited regions on a small creative-editing evaluation while supporting subsecond incremental updates.

Boxer: Robust Lifting of Open-World 2D Bounding Boxes to 3D

Boxer separates open-vocabulary 2D detection from 3D geometry estimation, lifts proposals with BoxerNet using optional sparse or dense depth, and fuses predictions into static scene-level boxes; with ground-truth 2D boxes and dense depth on CA-1M, it improves per-frame class-agnostic mAP from CuTR's 0.250 to 0.412.

DreamPartGen: Semantically Grounded Part-Level 3D Generation via Collaborative Latent Denoising

DreamPartGen couples part geometry and appearance with language-derived relational semantics through intra-part and inter-part synchronized denoising, reducing CD on PartRel3D from PartCrafter's 0.371 to 0.081 while producing editable compositions, although low part overlap alone does not establish physically correct assembly.

On Geometric Understanding and Learned Priors in Feed-forward 3D Reconstruction Models

Layer-wise fundamental-matrix probes, attention knockouts, and controlled scene perturbations show that DUSt3R, VGGT, and Depth Anything 3 develop geometric representations associated with correspondence matching, while VGGT also fills in occluded correspondences using priors, making geometry and learned priors complementary rather than mutually exclusive explanations.

RAG-3DSG: Enhancing 3D Scene Graphs with Re-Shot Guided Retrieval-Augmented Generation

RAG-3DSG diagnoses conflicting multi-view captions using virtual re-shots of object point clouds, then retrieves reliable neighboring objects to rectify uncertain nodes, improving SceneFun3D object R@10 from OpenFunGraph's 87.8% to 90.2% with GPT-4o.

ReconPhys: Reconstruct Appearance and Physical Attributes from Single Video

ReconPhys connects a frozen single-image 3D Gaussian reconstructor and a video-based physical parameter predictor to differentiable spring-mass simulation, learning simulation-ready assets from monocular free-fall videos through image reconstruction error and improving synthetic future-prediction PSNR from Spring-Gaus's 13.27 to 21.64 with reported feedforward inference under 1 second.

URoPE: Universal Relative Position Embedding across Geometric Spaces

URoPE lifts key image patches into 3D points along camera rays at fixed depth anchors, then reprojects them onto the query view image plane. This enables standard 2D RoPE to directly encode relative positional relationships across views or geometric spaces. It serves as a parameter-free, intrinsic-aware, coordinate-system-agnostic, and FlashAttention-compatible universal position embedding.

World Reconstruction From Inconsistent Views

The paper treats geometric drift in generated videos as observation deformation to remove, combining non-rigid point cloud alignment with deformation-aware 2D Gaussian Splatting (2DGS) optimization to reconstruct a single static world, achieving 79.29 in 3D consistency and 86.59 in photometric consistency across video models.

Any to Full: Prompting Depth Anything for Depth Completion in One Stage

Any2Full encodes sparse depth as scale prompts that guide geometrically consistent relative depth prediction inside frozen Depth Anything v2, then recovers metric scale in a single final fit, achieving an average rank of 2.3 across domain and missing-pattern evaluations and approximately 1.4 times the inference speed of PriorDA with the same DA-L backbone.

EventVGGT: Exploring Cross-Modal Distillation for Consistent Event-based Depth Estimation

EventVGGT transfers VGGT's multi-view geometric knowledge into event sequences through mixed-feature auxiliary predictions, spatio-temporal feature alignment, and depth-change distillation, reducing EventScape's mean absolute depth error at the 30 m cutoff from EventDAM's 2.30 m to 1.06 m while using only events at inference time.

GaussianGPT: Towards Autoregressive 3D Gaussian Scene Generation

GaussianGPT compresses 3D Gaussian scenes into discrete token sequences and employs a causal Transformer with 3D rotary position embeddings for step-by-step autoregressive token prediction. This introduces the first purely autoregressive paradigm for 3D Gaussian scene generation, achieving or outperforming state-of-the-art diffusion models in shape synthesis and scene generation while naturally supporting scene completion and outpainting.

Holo360D: A Large-Scale Real-World Dataset with Continuous Trajectories for Advancing Panoramic 3D Reconstruction and Beyond

Holo360D combines continuous handheld panoramic capture, offline geometric alignment, and human-assisted repair to build a dataset of 109,495 panoramas, showing that horizontal perspective decomposition and mesh-depth fine-tuning recover more coherent geometry with existing feed-forward models than direct panoramic input.

MonoArt: Progressive Structural Reasoning for Monocular Articulated 3D Reconstruction

MonoArt organizes single-image articulated reconstruction into geometry generation, part representation, dual-query motion decoding, and kinematic estimation, achieving an F-Score of 0.670 and joint type accuracy of 67.47% on the 46-category PartNet-Mobility setting, although its axis direction error is not the best.

Syn4D: A Multiview Synthetic 4D Dataset

Syn4D combines high-quality dynamic scene rendering with compact mesh-based tracking annotations, providing approximately 4.7K multiview clips and 1.4M frames and improving dense tracking APD on Soviet from 58.35 to 74.46 without changing the 4RC architecture.

TORA: Topological Representation Alignment for 3D Shape Assembly

During the training of a flow-matching-based 3D assembly model, TORA additionally aligns its feature representation with the "who-is-similar-to-whom" relational structure of a frozen, pre-trained 3D encoder (matching Gram matrices via CKA loss). This distills geometric interaction priors into the network, accelerating convergence by up to \(6.9\times\), significantly enhancing cross-domain robustness, and incurring zero extra cost during inference.

Fisheye3R: Adapting Unified 3D Feed-Forward Foundation Models to Fisheye Lenses

Fisheye3R equips frozen unified feed-forward 3D models with learnable calibration tokens and camera-type attention masks for native fisheye and mixed-camera inputs; adaptation using only unlabeled perspective images raises MapAnything's pose AUC on ScanNet++ fisheye testing from 0.519 to 0.766.

FLEG: Feed-Forward Language Embedded Gaussian Splatting from Any Views via Compact Semantic Representation

FLEG reconstructs language-embedded 3D Gaussians from any number (2 to 30+) of uncalibrated, unposed multi-view images in a single feed-forward pass: it trains without any 3D ground truth by distilling geometric pseudo-labels from a frozen VGGT and semantics from SAM2 plus CLIP, and it decouples the language embedding from dense geometric Gaussians into a sparse set of semantic Gaussians, preserving semantic fidelity with less than 5% of the embeddings.

GaussianLens: Localized High-Resolution Reconstruction via On-Demand Gaussian Densification

GaussianLens takes a low-resolution 3DGS reconstruction as a coarse scaffold, and for a region of interest (RoI) selected by the user with a 2D mask it takes sparse high-resolution images of that region as an extra input; a cross-modal network then predicts residual updates for the Gaussians inside the RoI and additionally sprouts one "pixel-guided Gaussian" per high-resolution pixel, turning localized high-fidelity reconstruction from "raise the global resolution" into one on-demand forward pass — 28.54 / 23.74 dB PSNR on RE10K 256→512 and DL3DV 256→1024, over 1 dB above comparable baselines, with 40% of the Gaussians a uniformly high-resolution model needs.

GENA3D: Generative Amodal 3D Modeling by Bridging 2D Priors and 3D Coherence

GENA3D proposes a generative amodal 3D modeling framework. Through View-Wise Cross-Attention (which parallelly fuses multi-view 2D completed features) and Stereo-Conditioned Cross-Attention (which injects MVS point clouds as geometric gating into attention logits), it achieves both generative diversity and geometric coherence under sparse, occluded inputs. It reaches an FID of 30.73 (4-view) on the GSO dataset, surpassing Amodal3R by 4.4, with a visible-region SSIM of 0.838.

OmniX: Any-view and Any-time 4D reconstruction via Feed-forward Trajectory Fields

OmniX separates per-frame geometry recovery from temporal motion prediction, generates trajectory transformation bases from a small set of dynamic tokens, and samples them into pixel-level trajectories, improving foreground APD3D from VDPM's 0.199 to 0.300 in the monocular, all-frames evaluation on its custom validation set.

OmniX: From Unified Panoramic Generation and Perception To Graphics-Ready 3D Scenes

OmniX turns a pre-trained 2D flow matching model (Flux.1-dev) into a unified "multi-condition in, multimodal out" panorama model: circular-translation-equivariant operators remove equirectangular seams, modality-specific LoRAs isolate each conditioning input, and one framework covers image-to-panorama generation, RGB-to-distance/normal/albedo/roughness/metallic perception, and masked completion; distance and material panoramas are then projected into a textured triangle mesh, yielding graphics-ready 3D scenes that import directly into Blender for relighting and physical simulation.

One Video, One World: Turning Monocular Video into Physical 4D Scenes

OVOW is the first training-free system that transforms monocular videos into instance-level, physically simulatable 4D mesh scenes. Through a four-stage pipeline consisting of VLM-guided scene decomposition, category-aware mesh reconstruction, iterative scale-pose recovery, and physically grounded assembly, it outputs watertight mesh instances without any fine-tuning, which can be directly imported into physical simulators like MuJoCo.

Predictive Photometric Uncertainty in Gaussian Splatting for Novel View Synthesis

Uncertainty is turned into a spherical-harmonic channel on each 3D Gaussian that is isomorphic to color; it is fitted to training-view residual maps by solving a linear least-squares problem built from frozen alpha-blending weights and regularized with a Bayesian prior, yielding per-pixel uncertainty maps for arbitrary novel views — AUSE(DSSIM) on Mip-NeRF360 drops from 0.495 to 0.214 and Pearson correlation rises from 0.160 to 0.547 at only 12.8% extra training time.

Towards Geometry-Grounded Dense Semantic Matching with VGGT Priors

The paper transfers VGGT from same-instance multiview reconstruction to cross-instance dense semantic matching through direct bidirectional sampling-grid prediction, synthetic dense supervision, and cycle-consistent training, improving correspondence accuracy, local structural preservation, and reliability estimation for invisible regions.

Volume Transformer: Revisiting Vanilla Transformers for 3D Scene Understanding

This paper ports a ViT-style vanilla Transformer encoder (volumetric cubic-patch tokenization + full global self-attention + 3D RoPE) straight into 3D scene understanding, compensates for the scarcity of current 3D supervision with a data-efficient recipe (strong augmentation, strong regularization, distillation from a convolutional teacher), and then scales supervision via multi-dataset joint training — setting new best results on ScanNet / ScanNet200 / ScanNet++ and nuScenes / Waymo / SemanticKITTI for both semantic and instance segmentation while being faster and more memory-efficient at inference than PTv3.

AerialMetric: Benchmarking and Adapting UAV Monocular Metric Depth Estimation in the Real World

This paper constructs a large-scale UAV aerial monocular metric depth benchmark, AerialMetric, consisting of four complementary subsets (52K real + 16K synthetic image-depth pairs). It systematically reveals the severe domain gap of existing SOTA models under aerial perspectives where δ1 is close to 0, and improves the aerial δ1 to over 84% through parameter-efficient fine-tuning of MoGe2 with LoRA, with almost no loss in ground generalization capability.

AirZoo: A Unified Large-Scale Dataset for Grounding Aerial Geometric 3D Vision

AirZoo constructs a million-scale synthetic aerial dataset. By utilizing an AirSim-Cesium-Unreal tri-simulation pipeline, it automatically renders UAV aerial sequences with pixel-level depth and 6-DoF georeferenced poses from Google 3D Tiles, covering 377 multi-weather trajectories across 22 countries. Fine-tuning existing SOTA models on three tasks—aerial image retrieval, cross-view matching, and multi-view 3D reconstruction—yields significant improvements, particularly enhancing zero-shot generalization capabilities in real-world scenes.

Auto3R: Automated 3D Reconstruction and Scanning via Data-driven Uncertainty Quantification

Auto3R learns uncertainty from the current 3D model's color and depth renders, selects the next scanning viewpoint through depth weighting and reliability reweighting, achieves 27.00 dB average PSNR on Objaverse with a final budget of 20 views, and extends view selection to continuous robotic scanning paths.

Drop-In Perceptual Optimization for 3D Gaussian Splatting

Treating the 2D distortion loss used in 3DGS training as the single design variable, this paper compares the original L1+SSIM loss, a composite L1+L2+MS-SSIM+LPIPS loss, and Wasserstein Distortion (WD), and picks the winner with a large-scale subjective study of 39,320 pairwise ratings: a lightly regularized WD variant, WD-R, replaces the loss and nothing else, yet is preferred by raters more than 2.3× over the original loss and 1.5× over Perceptual-GS, and it transfers unchanged to Mip-Splatting, Scaffold-GS, and 3DGS compression (≈50% bitrate savings).

E-M3RF: An Equivariant Multimodal 3D Re-assembly Framework

E-M3RF fuses rotation-consistent geometric features with pointwise color, predicts fragment poses through SE(3) flow matching, and penalizes interpenetration during training, reducing RePAIR rotation RMSE from GARF's 49.31 degrees to 37.91 degrees without leading on every metric of every generalization dataset.

Edit3r: Instant 3D Scene Editing from Sparse Unposed Images

Edit3r turns 3D scene editing into a single forward pass of sparse unposed Gaussian reconstruction: it trains on a SAM2-based recoloring pretext that is view-consistent by construction, and deliberately pairs one recolored view with one raw view so the feed-forward backbone learns to fuse conflicting per-view 2D edit proposals into a geometrically coherent edited 3D Gaussian scene, producing results in about 0.5 seconds at inference.

FILT3R: Latent State Adaptive Kalman Filter for Streaming 3D Reconstruction

FILT3R recasts the recurrent state update of a streaming reconstruction model as state estimation in token space: process noise is estimated online from EMA-normalized temporal drift while measurement noise is pinned to a single scalar, so each token's gain automatically shrinks as confidence accumulates during stable stretches and rises when the scene genuinely changes — markedly reducing long-horizon drift and catastrophic forgetting with a frozen backbone and no training at all.

FoundationGeo: Learning Spatial Pixel-Wise Fields for Monocular Metric Geometry

FoundationGeo learns a strong relative point map, recovers physical scale through pixel-wise scale fields and ray-direction correction, and repairs camera-distribution gaps with focal-diverse synthetic data, reaching average AbsRel of 14.8% and \(\delta_1\) of 80.8% across seven zero-shot metric-depth benchmarks.

Free-Range Gaussians: Non-Grid-Aligned Generative 3D Gaussian Reconstruction

Free-Range Gaussians directly generates 8K Gaussians untied to pixel or voxel grids from four posed images, combining hierarchical pairing and reconstruction guidance to improve unseen-region completion, reaching 28.08 dB PSNR and 27.69 FID on partially observed GSO; its iterative generation is not single-pass reconstruction.

From Perspective to Fisheye Depth Estimation and Open-Vocabulary Segmentation

The paper introduces Distortion Extenders (DEX), a set of lightweight learnable modulators inserted into every layer of a frozen backbone, which pull the feature distribution of fisheye images back toward the perspective feature distribution so that depth estimators and open-vocabulary segmentation models pretrained only on perspective data transfer zero-shot to fisheye cameras; training needs neither ground truth nor real fisheye images, and inference needs neither calibration nor re-projection. DEX improves RMSE/δ1 and mIoU on three real fisheye datasets (ScanNet++, KITTI-360, WoodScape), and its activations can be decoded back into KB distortion coefficients for camera calibration.

FUSE: A Flow-based Mapping Between Shapes

Each shape is represented as an invertible continuous flow from a shared unit-Gaussian anchor (trained with flow matching); composing the source shape's inverse flow with the target shape's forward flow yields a pointwise map between them — no paired training and no test-time optimization, plus cross-representation matching across meshes, point clouds, SDFs and tetrahedral meshes, reaching the lowest or near-lowest correspondence error on landmark-supervised benchmarks such as FAUST and SMAL.

GAINS: Gaussian-based Inverse Rendering from Sparse Multi-View Captures

GAINS augments a two-stage 2D Gaussian Splatting inverse-rendering pipeline with five systematically integrated foundation-model priors (monocular depth/normal, geometry-stage SDS, SAM subpart segmentation, monocular intrinsic image decomposition, and multi-illuminated SDS), disentangling geometry, materials and environment lighting from only 4-32 input views and reaching relighting PSNR of 28.16 on Synthetic4Relight (Ref-Gaussian: 24.29) and 22.29 on Shiny Blender (Ref-Gaussian: 17.26).

Geometry-Aware Single-Image 4D Synthesis via Dense Trajectory Generation

MoGe4D estimates geometry from one image, generates geometry-conditioned dense motion trajectories with a diffusion model, and completes novel-view videos, improving imaging quality and trajectory error under complex camera motion without leading every video consistency or dynamics metric.

GeoWorld: Providing Full-frame Geometry Features to Facilitate 3D Scene Generation

GeoWorld completes a coarse video along a camera trajectory, then uses full-frame VGGT geometry as conditioning and supervision for a second video stage, reaching 17.28 dB novel-view PSNR on RealEstate10K and 16.64 dB after reconstruction into a 3DGS scene.

GRAFT: Geometric Refinement and Fitting Transformer for Human Scene Reconstruction

GRAFT is a learned "human-scene interaction prior" that amortizes the optimization-based physical plausibility fitting process into a feed-forward Transformer. By encoding the relative human-scene relationship into 24 compact tokens via geometric probes, it predicts "interaction gradients" in a recurrent loop to correct floating and penetration. Ultimately, it achieves optimization-level interaction quality with a feed-forward speed of 0.2s and can serve as a plug-and-play prior to enhance any feed-forward method.

Hierarchical and Holistic Open-Vocabulary Functional 3D Scene Graphs for Indoor Spaces

The paper extends functional 3D scene graphs to tabletop manipulable objects and internal object hierarchies, using 2D visual anchoring, multi-cue node association, temporal edge optimization, and global hierarchy shaping to outperform OpenFunGraph Hierarchy by 13.9 and 20.5 percentage points in overall node and triplet retrieval on extended FunGraph3D.

HSImul3R: Physics-in-the-Loop Reconstruction of Simulation-Ready Human–Scene Interactions

HSImul3R turns the simulator into a reconstruction supervisor, using scene-targeted reinforcement learning to refine human motion and direct simulation reward optimization to refine object geometry, raising easy-case HSIBench interaction stability from HSfM's 10.52% to 53.68%, although hard-case stability remains only 13.92%.

MAGiSt3R: Multi-Agent Feed-forward 3D Reconstruction from Monocular RGB Videos

MAGiSt3R predicts local geometry with VGGT, uses a shared MAGMA network for both intra-agent and inter-agent submap registration, and corrects drift through pose graph optimization, enabling multi-agent indoor reconstruction from uncalibrated RGB videos with mean multi-agent reconstruction errors of 5.37/3.36 in Acc./Comp. on ReplicaMultiagent and a two-agent runtime of 9 FPS.

Masked BRep Autoencoder via Hierarchical Graph Transformer

The method preserves CAD boundary-representation topology while masking raw face and edge geometry and attributes, reconstructing them through dual-resolution cross-scale attention and local message passing to learn transferable representations that reach 88.75% face-level accuracy on MFInstSeg with 0.1% labeled data.

Matryoshka Gaussian Splatting

MGS dynamically orders Gaussians by opacity and jointly trains a random-length prefix and the full set, enabling fine-grained budget truncation from one scene model with 28.20 dB full-budget PSNR and a quality–FPS AUC of 64.81 on MipNeRF 360, at roughly twice the backbone training cost.

MoCam: Unified Novel View Synthesis via Structured Denoising Dynamics

MoCam hands the two conditioning signals — a point-cloud scaffold rendered along the target trajectory and the source-view video — to a video diffusion model on a relay along the denoising timeline: the scaffold anchors geometry and motion early, then a switch hands over to the source video for error correction and detail refinement, unifying single-image 3D reconstruction and video 4D re-camera in one pipeline.

Neural Harmonic Textures for High-Quality Primitive Based Neural Reconstruction

The paper attaches learnable feature vectors to the virtual tetrahedron bounding each primitive (Gaussian or triangle), barycentrically interpolates them at ray–primitive intersections, applies sin/cos periodic activations, and then alpha-composites along the ray — turning the blend itself into a harmonic decomposition — so that a single image-space pass of a shallow MLP decodes the final color; it surpasses 3DGS, 2DGS, and Spherical Voronoi at 140+ FPS on MipNeRF360, Tanks & Temples, and Deep Blending (28.74 PSNR on MipNeRF360, 25.68 on Tanks & Temples) and plugs into 2DGS and Triangle Splatting unchanged.

Point2Pose: Occlusion-Recovering 6D Pose Tracking and 3D Reconstruction for Multiple Unknown Objects Via 2D Point Trackers

Point2Pose turns a long-range 2D point tracker into a persistent data-association engine: a set of query points is scattered on each object, lifted to 3D with depth and accumulated into an object-centric keypoint map, and the pose follows from frame-to-map registration. This lets it track multiple unknown rigid objects simultaneously without any CAD model and recover immediately once an object reappears after complete occlusion — at the cost of slightly lower single-object ADD and reconstruction accuracy than BundleSDF.

Pseudo-Stereo Inputs: A Solution to the Occlusion Challenge in Self-Supervised Stereo Matching

The paper replaces one view of the input pair with a pseudo-view rendered by the model currently being trained, decoupling the network's inputs from the photometric feedback so that occlusion falls on either side of an occluder with equal probability; combined with a rendering-derived occlusion mask and a fully pseudo-stereo input strategy, it lowers D1-all on the KITTI 2015 test set from the previous best of 6.11% to 4.39% and cuts the D1 error inside occluded regions from 20.82% to 9.05%.

SeeClear: Reliable Transparent Object Depth Estimation via Generative Opacification

SeeClear uses diffusion to replace transparent objects with opaque appearances that retain contours and surface shading, then applies an unmodified depth model, reducing DA3 RMSE from 24.03 to 21.45 mm on transparent-object regions in ClearGrasp.

SKEL-CF: Coarse-to-Fine Biomechanical Skeleton and Surface Mesh Recovery

SKEL-CF proposes a coarse-to-fine framework to recover biomechanical skeletons and surface meshes from a single RGB image. Under this framework, the model is first trained on the high-quality HMR-SKEL dataset, followed by explicit camera intrinsic modeling to eliminate depth ambiguity. Finally, through an encoder coarse prediction + decoder layer-wise iterative refinement strategy, it significantly outperforms the previous SKEL method HSMR on challenging datasets such as MOYO (reducing MPJPE by 18.6%), while achieving comparable or even superior numerical accuracy to SMPL-based methods, and outputting anatomically more plausible poses.

TACO-Net: Topological Signatures Triumph in 3D Object Classification

TACO-Net converts point clouds into fixed topological statistics from multiple grayscale voxel volumes and trains a lightweight 1D CNN to combine them, reaching 99.05% OA on ModelNet40, while the full pipeline including topological feature generation processes approximately 5.3 point clouds/second.

WAFT-Stereo: Warping-Alone Field Transforms for Stereo Matching

WAFT-Stereo replaces cost volumes with a single disparity-classification initialization followed by recurrent regression driven by high-resolution feature warping, achieving 0.89% ETH3D BP-0.5-noc with synthetic-only training and 106 ms latency for 540p inputs on an L40 with DAv2-L and 5 iterations.

WorldMesh: Generating Navigable Multi-Room 3D Scenes via Mesh-Conditioned Image Diffusion

WorldMesh first builds a multi-room mesh carrying building structure and reconstructed objects, then feeds progressively accumulated wall textures and depth as conditioning to image generation, and finally reconstructs the scene with mesh-depth-regularized 3D Gaussian Splatting (3DGS), achieving a 96.2% average preference over baselines in a 31-participant perceptual study; its core is a persistent geometric constraint rather than merely generating a video that looks coherent.

3D-LENS: A 3D Lifting-based Elevated Novel-view Synthesis method for Single-View Aerial-Ground Re-Identification

3D-LENS lifts images from a single real viewpoint domain into fixed textured meshes, renders missing aerial or ground views, and combines appearance alignment with robust mixed-domain training, reaching 37.8% mAP on AG-ReID.v2 CCTV-to-aerial retrieval under aerial-only training, 21.2 percentage points above the strongest prior method in that column, PASS, without real target-view training samples.

Axolotl3D: a Unified Framework for Faithful 3D Shape Completion

Axolotl3D unifies images, visibility masks, cameras, and partial points as conditions for 3D generation, using geometric anchors to achieve a 0.9046 F-score on occluded single-view Toys4K while also supporting local shape edits that preserve unedited regions.

Beyond Prompts: Unconditional 3D Inversion for Out-of-Distribution Shapes

The paper shows that native text-to-3D generative models (TRELLIS) fall into a "sink trap" where they become insensitive to out-of-distribution text conditioning, so it replaces the source prompt used for inversion with an empty prompt and anchors the shape purely in the unconditional prior — cutting the inversion reconstruction error on non-rigid shapes from L1 17.75 to 5.40 and unlocking the first 3D editing pipeline driven only by a target prompt, with no masks and no 2D priors (SigLIP 0.1469, about 1.8x the strongest baseline, averaging 9 seconds per edit).

BrepLLM: Enabling Large Language Models to Understand Boundary Representations

BrepLLM is the first framework that enables large language models to directly parse and reason over native B-rep (Boundary Representation) CAD models. By performing adaptive UV sampling to convert B-rep models into face-edge topology graphs, and using a hierarchical BrepEncoder to extract decoupled face, edge, and topology features, followed by CLIP contrastive pretraining and two-stage progressive LLM fine-tuning, it comprehensively outperforms point cloud- and image-based baseline models on 3D object description and classification tasks.

CanoVerse: 3D Object Scalable Canonicalization and Dataset for Generation and Pose

CanoVerse replaces continuous 3D rotation annotation with human selection among five candidate poses, creating a canonical dataset of 320K objects across 1,156 categories that brings VI-Net to 20.18% Acc@10° on OmniObject3D and substantially reduces pose variation in image-to-3D generation.

City-Level 3D Surface Reconstruction with Viewpoint Orientation Partitioning and Scene Completion

The method groups camera views that support each other's reconstruction, selectively completes the initial point cloud using keypoint distribution entropy, and optimizes Gaussian representations with multi-view consistency, increasing MatrixCity surface reconstruction F1 from CityGS-X's 0.581 to 0.803.

DDStereo: Efficient Dual Decoder Transformers for Stereo 3D Road Anomaly Detection

DDStereo splits category-agnostic localization from semantic classification into two decoders — a 3D main branch that regresses size, orientation and known-class scores from left-view texture features, and a foreground branch that only performs binary obstacle classification on stereo disparity features — with both branches sharing one set of object queries for instance alignment, so that the anomaly score is simply the foreground confidence minus the maximum known-class probability; it reaches 78.09% OoD AP3D on AR-OoD (+4.40 over S3AD) while running at 24 ms with 19.6 M parameters on KITTI, matching monocular detectors in speed.

DefenseSplat: Enhancing the Robustness of 3D Gaussian Splatting via Frequency-Aware Filtering

DefenseSplat significantly improves the robustness of 3DGS against adversarial attacks without clean ground-truth supervision. It decomposes input images using the wavelet transform and filters out high-frequency subbands to suppress adversarial perturbations, while introducing a ReLU-based scale regularization loss to constrain the generation of elongated Gaussians.

Dense Dynamic Scene Reconstruction and Camera Pose Estimation from Multi-View Videos

This paper proposes a two-stage optimization framework for the setting in which several freely moving cameras film the same dynamic scene: a feed-forward reconstruction model (VGGT) provides wide-baseline initialization and a shared scale anchor, a spatio-temporal connection graph unifies intra-camera temporal, same-timestamp inter-camera, and cross-camera historical constraints into a joint multi-camera bundle adjustment, and a final stage uses wide-baseline dense optical flow for frame-level scale alignment followed by alternating pose and per-pixel depth refinement. It outperforms COLMAP/GLOMAP and VGGT/Fast3R/FastVGGT/CUT3R in pose accuracy, scene consistency, and GPU memory, and releases the motion-capture-grounded MultiCamRobolab dataset.

DreamCAD: Scaling Multi-modal CAD Generation using Differentiable Parametric Surfaces

DreamCAD uses differentiable parametric surfaces with shared boundaries to learn from point supervision on large-scale meshes, then supports image, point-cloud, and image-mediated text conditioning through coarse-to-fine flow matching, reducing ABC point-to-CAD CD from Cadrille's 2.98 to 0.93 (both displayed as raw CD multiplied by \(10^3\)), while leaving complete CAD topology to an additional recovery stage.

EditVerse3D: High-Quality 3D Object Editing with Region-Aware Learning

EditVerse3D turns TRELLIS into a supervised editor conditioned on an original object, a coarse 3D bounding box, and a local reference image, combining coarse-input training with region-aware loss reweighting to improve whole-output PSNR from Repaint's 27.44 to 28.67 on addition while reducing the need for precise masks and fully edited views.

Face Anything: 4D Face Reconstruction from Any Image Sequence

The paper proposes Face Anything, unifying 4D face reconstruction and tracking as canonical coordinate prediction—the network jointly estimates depth and canonical face coordinates from arbitrary image sequences, and dense correspondences are obtained directly through nearest-neighbor search in canonical space without explicitly modeling inter-frame motion. It outperforms existing methods across depth estimation (16% lower RMSE), correspondence accuracy (3× improvement), and inference efficiency (32× speedup).

Fourier Splatting: Generalized Fourier encoded primitives for scalable radiance fields

This paper replaces the fixed disc boundary of a planar surfel with a closed curve defined by a Fourier series that can be truncated at any time, so that a single trained model carries a continuously adjustable level of detail: simply transmitting fewer Fourier coefficients yields a lower-bitrate, lower-detail render without retraining or pruning, giving the best PSNR/SSIM among planar-primitive methods on Mip-NeRF 360.

From Blobs to Spokes: High-Fidelity Surface Reconstruction via Oriented Gaussians

The paper attaches a learnable oriented normal to every 3D Gaussian, reinterpreting it from an orientation-free symmetric blob into a stochastic surface element that attenuates only in its outward half-space; this restores the reciprocal attenuation assumed by Objects as Volumes-so occupancy and normal fields follow in closed form-and combined with a normal alignment loss, normal-aware densification and Primal Adaptive Meshing it sets a new state of the art on DTU and Tanks and Temples with only two pivots per Gaussian, far lighter meshes than concurrent work, and stable recovery of thin structures such as bicycle spokes.

Gaussian Belief Propagation Network for Depth Completion

GBPN reformulates depth completion as "performing Gaussian belief propagation on a Markov Random Field (MRF) dynamically constructed by a network." Instead of directly regressing depth, the network learns the MRF's potential functions, edge weights, and non-local edge structures, and then uses differentiable Gaussian BP to iteratively infer the dense depth distribution (treating the mean as depth and precision as confidence). It achieves SOTA performance on NYUv2 and KITTI, exhibiting robustness under extremely sparse inputs that far surpasses pure regression networks.

Gaussians on Fire: High-Frequency Reconstruction of Flames

The paper redefines fire as a population of short-lived volumetric emission carriers rather than persistent scene primitives: each 3D Gaussian carries a lifetime and a linear velocity, is initialized from a 3D flow field fused from three-view dense optical flow, and is captured with an ESP32 + LED sub-frame synchronization rig, so that three consumer cameras suffice to reconstruct a temporally coherent high-frequency 4D flame representation that beats 4DGS, Grid4D and FreeTimeGS on flame-region PSNR/SSIM/LPIPS, depth consistency, and motion-field similarity on synthetic scenes.

Generalizable Neural Reconstruction of High-Fidelity Surfaces via Sparse Volumetric Representations

SVRecon replaces the expensive dense feature volume in generalizable neural surface reconstruction with a nested two-stage design — low-resolution occupancy prediction followed by a high-resolution sparse feature volume — pushing feed-forward reconstruction from 128³ to 512³ within 32GB of VRAM and delivering the smoothest, most normal-consistent surfaces on sparse-view DTU.

HandSCS: Structural Coordinate Space for Animatable Hand Gaussian Splatting

HandSCS assigns skeleton-relative structural coordinates to hand Gaussians, combines them with per-Gaussian residual embeddings and cross-pose matching, and achieves 32.91 dB PSNR and 0.0291 LPIPS for novel-pose rendering on InterHand2.6M test/Capture0, with some SSIM results still below competing baselines.

Heat Kernel Textures -- the Geodesic Gaussians That Do Not Splat

HKTex places learnable anisotropic heat kernels directly on triangle-mesh surfaces and combines surface-constrained optimization with differentiable ray tracing to replace UV texture storage, recovering detail with a smaller reported storage footprint but without the fast rasterization advantage of Gaussian Splatting.

HiReFF: High-Resolution Feedforward Human Reconstruction from Uncalibrated Sparse-View Video

HiReFF builds on a VGGT backbone with Gaussian, mask, and high-resolution side-tuning branches to feed-forward reconstruct human 3D Gaussians from four uncalibrated 90°-spaced videos, rendering 360° dynamic humans at 2072×2072 and streaming reconstructions at 3.01 FPS on a single RTX 4090.

Improving Sparse-View 3DGS Generalization via Flat Minima Optimization

This work addresses overfitting in sparse-view 3DGS from the perspective of flat minima optimization: through scale-adaptive position perturbation (SAP) + random application + magnitude scheduling + periodic reinitialization of non-position parameters, it drives 3DGS to converge to flatter regions of the loss landscape without changing the network architecture, achieving SOTA or near-SOTA rendering quality under sparse-view settings on LLFF and Mip-NeRF360.

Interact3D: Compositional 3D Generation of Interactive Objects

Interact3D proposes a training-free “generate-then-compose” framework. It leverages the spatial priors of 2D/3D generative models to formulate 3D compositional generation as a structured registration problem. Through a two-stage composition pipeline—global-to-local geometric alignment to anchor the main object, followed by SDF collision-aware optimization to place the remaining objects—and a VLM-driven agent closed-loop for iterative correction, it automatically generates physically plausible and collision-free interactive 3D compositional scenes from a single image and text prompt, while releasing a dataset of 8,000+ interactive 3D pairs.

Interaction-Aware 4D Gaussian Splatting for Dynamic Hand-Object Interaction Reconstruction

Without any object shape, category, or pose prior, this paper drives 4D Gaussian Splatting with three decoupled implicit fields (hand, object, background), augments each Gaussian with a learnable weight \(w\) and radius \(o\) to model occlusion and edge sharpness, conditions the object deformation field on key-frame hand positions, and combines explicit hand / object-rotation / interaction regularizers with a five-phase progressive optimization to surpass 4DGS, Deform3DGS, SC-GS and the specialized HOLD / BIGS on HOI4D and HO3D (PSNR 30.32 dB on HOI4D translation scenes, 3.99 dB above the strongest baseline).

IRIS: Intersection-aware Ray-based Implicit Editable Scenes

IRIS attaches neural features to movable Gaussians and replaces dense sampling plus per-query 3D neighbor searches with analytical ray sampling and local feature aggregation, achieving 32.65 dB PSNR at 6.038 FPS on NeRF Synthetic; its principal advantage is efficiency over earlier editable neural fields, not real-time rendering in every scene.

Know3D: Prompting 3D Generation with Knowledge from Vision-Language Models

Know3D injects intermediate backside-generation features from Qwen-Image-Edit into TRELLIS.2 to impose language constraints on unseen geometry, achieving ULIP 0.2174 and Uni3D 0.3518 on HY3D-Bench-test, while backside instruction following is demonstrated primarily through qualitative examples.

LinStereo: Linear-Complexity Global Attention for Multi-Scale Iterative Stereo Matching

LinStereo replaces the ConvGRU update operator of iterative stereo matching — which only ever reasons over a local window — with a Position-Aware Linear Attention (PALA) operator that aggregates global context over the whole feature map at \(O(N)\) cost per step, and pairs it with per-scale Hierarchical Semantic Cost Volumes (HSCV) and a monocular Depth Prior Initialization (DPI); trained on SceneFlow alone it reaches state-of-the-art-level accuracy on standard benchmarks and the best overall accuracy on TartanAir-UW / SQUID without any underwater training data.

MAC-Splat: Multi-Attribute Consistency for High-Fidelity Sparse-View Reconstruction

MAC-Splat connects 3D Gaussians predicted from different views using semantically enriched 2D matches and jointly regularizes position, shape, and appearance, reaching 21.34 dB PSNR on the ScanNet++ Very Wide split, 5.18 dB above Splatt3R.

Markov-Renewal Single-Photon LiDAR Simulator

MaRS formulates photon registration under SPAD dead time as a Markov-renewal process (MRP), analytically predicting each pixel's photon-count distribution (Gaussian mean and variance) and its temporal PDF; delay invariance plus a precomputed lookup table then synthesize a whole H×W×n_b histogram cube in one parallel sampling pass — statistically indistinguishable from per-photon rejection-sampling gold standard, yet generating a 128×128×1024 cube in 1.01 s versus 40.8 s (>40×).

MLP Splatting: Object-Centric Neural Fields

MLP-Splatting replaces each scene primitive's role as "a Gaussian ellipsoid" with "a compact MLP that shares no parameters and only acts inside its own anisotropic Gaussian support," compositing images through tile-sorted sparse volume rendering; under RGB supervision alone these neural primitives spontaneously specialize to objects or object parts, so object-level editing needs no segmentation masks, and on Replica/ScanNet the method matches or beats Feature-3DGS at roughly 1/7 the memory and 5x the speed.

Monocular Avatar Reconstruction via Cascaded Diffusion Priors and UV-Space Differentiable Shading

Using a unified pretrained diffusion backbone to pipeline three sub-tasks: "texture completion \(\rightarrow\) illumination homogenization \(\rightarrow\) material decomposition", with each sub-task attaching only a single LoRA, combined with a UV-space differentiable BRDF rendering loss, high-fidelity and relightable 3D facial PBR avatars at 4K resolution can be reconstructed from a single in-the-wild image using fewer than 100 real 3D scans.

Monte Carlo Energy Aggregation for Mobile 3D Gaussian Splatting

Flux-GS compresses 3rd-order Spherical Harmonics (SH) into a 1st-order representation via Monte Carlo specular energy aggregation. Complemented by an attribute-conditioned enhancement module with zero inference overhead and a multi-view alpha-weighted densification/pruning strategy, it achieves real-time, high-fidelity novel view synthesis at 130+ FPS on a mobile Snapdragon 8 Gen 3. It also compresses storage to 2-5 MB and reduces training time from Mobile-GS's 80+ minutes to 11 minutes.

MotionAnymesh: Physics-Grounded Articulation for Simulation-Ready Digital Twins

MotionAnymesh grounds VLM part grouping with SP4D kinematic priors, then estimates joints via a two-stage "contact-interface geometric initialization + physics-constrained trajectory optimization" scheme, turning a single static mesh zero-shot into a collision-free URDF digital twin with joint limits — reaching 0.86 segmentation mIoU and 87% physical executability in SAPIEN (versus 46% for the best prior method).

Multi-View Foundation Models

The paper inserts multi-view adapters into 2D foundation models to jointly process images of a static scene, using geometric correspondence supervision and semantic regularization to reduce DINOv2's ScanNet++ localization error from 0.1029 to 0.0247 while retaining 0.9376 cosine similarity to the original features.

OrthoTrack: Continuous 6-DoF UAV Trajectory Estimation Anchored in Public Orthophotos

OrthoTrack proposes a training-free, 6-DoF continuous UAV trajectory estimation system. By leveraging public orthophotos (DOP) and Digital Surface Models (DSM) using an alternating strategy of "keyframe dense matching anchoring + inter-frame optical flow propagation", it achieves sub-meter absolute metric pose estimation at real-time frame rates without requiring GPS, IMUs, or post-processing alignment. It also introduces MovingDrone, a large-scale multimodal benchmark.

Pix2NPHM: Learning to Regress NPHM Reconstructions From a Single Image

Pix2NPHM uses a geometrically pretrained, face-specific ViT to regress the disentangled identity and expression latents of MonoNPHM directly from a single image; trained on a self-curated set of 102K NPHM registrations plus normal-based 2D self-supervision, it outperforms all public single-image face reconstruction methods on NeRSemble SVFR and NoW while running at roughly 8fps on an RTX3080.

PixGS: Pixel-Space Diffusion for Direct 3D Gaussian Splat Generation

PixGS lays out every 3D Gaussian's attributes as a multi-view "attribute image" and denoises that attribute tensor directly with a pre-trained pixel-space diffusion model (PixNerd, ~45M-image priors) under a flow-matching objective, combining 2x super-resolution rendering supervision, depth/normal and multi-scale LoG regularization, and a three-phase training curriculum to produce a 3DGS asset in about one second on a single A100 for either text or image conditions — beating DiffSplat, TRELLIS, GaussianCube, LGM and DreamGaussian on both T3Bench and GSO.

RAGA: Real Time Ray Traced Gaussian Shadow Casting for 3DGS Avatar-Scene Interaction

RAGA proposes a ray tracing method for shadow computation completely within the Gaussian space—by computing precise Gaussian line integrals and normalizing them along each shadow ray to obtain a thickness factor describing how much of the occluding volume the light "passes through", which drives the shadow transmittance; meanwhile, it introduces an avatar proxy to eliminate temporal flickering in animations, achieving interactive shadow casting at ~50 FPS on NVIDIA OptiX.

RayMap3R: Inference-Time RayMap for Dynamic 3D Reconstruction

RayMap3R observes that a streaming reconstruction model trained with RayMap exhibits a static-scene bias when queried with RayMap features alone, and turns that bias into a training-free dynamic detector: the per-pixel depth discrepancy between a main branch and a RayMap-only branch gates memory-state updates, while reset metric alignment and state-aware smoothing keep long trajectories stable, pushing pose, depth, and reconstruction accuracy on dynamic scenes to the front of streaming methods with no training and constant memory.

Rotate Your Character: Revisiting Video Diffusion Models for High-Quality 3D Character Generation

RCM converts single or multiple images of characters in complex poses into high-resolution orbital videos in canonical A/T poses, combining video priors, per-pixel camera conditioning, and three-stage training for novel view synthesis before a separate reconstruction model produces meshes.

SGMatch: Semantic-Guided Non-Rigid Shape Matching with Flow Regularization

SGMatch fuses semantic features obtained by rendering shapes from multiple views, encoding them with DINOv2, and back-projecting them onto the surface into geometric descriptors via channel-wise gating plus mesh-neighborhood-restricted cross-attention, and adapts conditional flow matching (CFM) into a time-conditioned regularizer on vertex-level feature transport that smooths dense correspondence recovery, reaching the best results on strongly non-isometric (SMAL geodesic error 2.5) and topology-noisy (TOPKIDS 2.9) benchmarks.

ShellMaker: Language-Guided Exterior Completion under Structural Constraints

ShellMaker proposes a language-guided building exterior completion framework: given a building scaffold (wall + door/window opening layouts output by indoor generators/CityGML/CAD) and a style prompt, it produces a complete PBR exterior facade mesh through four modules: parametric roof generation, LLM part-aware prompt refinement, compatibility-aware joint wall-roof texture retrieval, and geometry-aware assembly. While strictly preserving footprints and opening layouts, it matches any architectural style, achieving a footprint IoU of 0.992 and an opening center error of only 0.091.

SK-Adapter: Skeleton-Based Structural Control for Native 3D Generation

SK-Adapter proposes 3D skeletons as a first-class control signal, injecting them into a frozen 3D flow matching generative backbone (Trellis) via a lightweight adapter network (GRPE topological encoding + skeletal cross-attention). This approach achieves precise structural control in native 3D space while preserving pre-trained generative priors, and is accompanied by the construction of Objaverse-TMS, a dataset containing 24k text-mesh-skeleton triplets.

StoryBlender: Inter-Shot Consistent and Editable 3D Storyboard with Spatial-temporal Dynamics

StoryBlender recasts "script to storyboard" from per-frame pixel sampling into hierarchical multi-agent planning inside a 3D engine: a four-layer continuity memory graph decouples global assets from shot-local variables, canonical assets are materialized as meshes in a unified coordinate space, and a reflection loop driven by engine physics checks plus VLM critiques solves layout, camera, and lighting — yielding both inter-shot consistency (CIDS Self 0.917 / Cross 0.803, OCCM 78.16) and non-destructive editability on CineBoard3D.

StructSplat: Generalizable 3D Gaussian Splatting from Uncalibrated Sparse Views

StructSplat assigns geometry, semantics, and texture to separate encoding paths and aligns cameras without leaking target-image features into Gaussian prediction, achieving 28.045 dB PSNR on DL3DV with two source views, no camera-parameter inputs, and no post-optimization.

SuperFlex: Deformable Superquadrics for Point Cloud Decomposition

SuperFlex learns bendable and taperable superquadrics with joint volumetric and surface losses, then supervises occluded inputs with optimized complete decompositions, raising feed-forward ShapeNet IoU from SuperDec's 0.59 to 0.72 while retaining a compact representation of 5.64 primitives on average.

SuperVoxelGPT: Adaptive and Ordered 3D Tokenization for Autoregressive Shape Generation

SuperVoxelGPT partitions and orders supervoxels using predicted geometric saliency, then generates local geometry tokens with Qwen2.5-0.5B and Jacobi decoding, achieving geometry quality close to TRELLIS2 with an average of 1,048 shape tokens and 4.60 seconds in image-to-3D tests.

TAIHRI: Task-Aware 3D Human Keypoints Localization for Close-Range Human-Robot Interaction

TAIHRI recasts task-relevant 3D human localization under a robot's egocentric camera as next-token prediction in a vision-language model: the interaction space is quantized into voxel tokens, the model reasons over 2D keypoints before inferring depth, and a pose-aware reward drives GRPO reinforcement fine-tuning — cutting global MPJPE on upper-body keypoints from 124.91 mm to 93.83 mm on Harmony4D and EgoBody close-range settings.

TanGO: Training-Free 3D Editing via Tangent-Space Guidance and Optimization

TanGO casts text-driven editing of a VecSet-based flow-matching 3D generative model as an instantaneous one-step optimal control problem solved at every ODE step: the per-token steering gain is determined by the directional mismatch between source- and target-conditioned velocity fields (a vMF negative log-likelihood, i.e. \(1-\cos\theta_i\)), and the closed-form solution pushes the tokens that must change while suppressing those that must be preserved — no masks, no fine-tuning, no gradient-based optimization. On the authors' TanGOEdit benchmark it reaches 0.2301 CLIP-T, 0.7679 CLIP-I and 0.6510 DINO-I, beating FlowEdit and AnchorFlow, at roughly 28.7s per edit.

TouchAnything: Diffusion-Guided 3D Reconstruction from Sparse Robot Touches

TouchAnything combines local geometry from sparse touches with a class-level shape prior from a frozen 2D diffusion model, reconstructing 3D objects through coarse-to-fine optimization without task-specific diffusion training and achieving lower EMD across six simulated categories with 20 touches.

TreeSRNF: Square-Root Normal Fields for Generative Modelling of the Geometric and Structural Variability in Tree-like 3D Objects

This paper generalizes the square-root normal field (SRNF) representation from fixed-topology genus-0 surfaces to tree-like 3D objects whose branching structure may change — each branch is parameterized as a surface over a unit cylinder and the branches are assembled recursively into a hierarchical tree — and then performs registration, geodesics and statistics in the resulting TreeSRNF tree-shape space under a three-term elastic metric (parent-branch bending/stretching + bifurcation position + subtree geometry), generating new 3D trees by sampling a multivariate Gaussian fitted in an analytically invertible tangent space.

TriFlow: Generating Artist-Like 3D Mesh Topology via Nearest-Vertex Vector Fields

TriFlow represents mesh topology as a nearest-vertex vector field (NVF) over the surface — every surface point points to the vertex with the largest barycentric weight in its local triangle — generates this field from an input SDF with conditional latent flow matching, and recovers the mesh through watershed clustering plus constrained QEM simplification. This bypasses the slow inference and error accumulation of autoregressive sequence modeling: on Objaverse the Chamfer Distance drops from 0.98 (best prior learned method) to 0.12, inference is about 8x faster, and the resulting topology wins across the board in both VLM and human preference studies.

TRiGS: Temporal Rigid-Body Motion for Scalable 4D Gaussian Splatting

TRiGS models each Gaussian primitive's motion as a continuous rigid-body trajectory in se(3) — a coupled exponential map yields rotation and translation jointly, a quadratic Bézier residual inside the visibility window captures nonlinearity, and a per-primitive local anchor serves as an independent center of rotation — reaching the best rendering quality on the 600/900/1200-frame SelfCap sequences and on N3V under a constant budget of 0.5M primitives and 160 MB.

UniQueR: Unified Query-based Feedforward 3D Reconstruction

UniQueR reformulates feedforward 3D reconstruction from per-pixel prediction into inference over a sparse set of 3D queries: 4096 learnable anchors with explicit 3D coordinates interact with multi-view features in a shared scene frame and each spawns 64 Gaussians, so a single forward pass turns unposed images into camera poses, point maps and renderable Gaussians — beating feedforward baselines on Mip-NeRF 360 / VR-NeRF in both rendering quality and depth accuracy with roughly 1/15 the number of primitives.

Video Generative Models as Geometry Learner

GeoNeXt treats depth and normals from one RGB image as subsequent video-model "frames" and jointly reconstructs appearance and geometry, reducing KITTI depth AbsRel from unified generative baseline GeoWizard's 14.4 to 8.2 with only 59K synthetic training samples, although it does not outperform specialist models on every metric.

Wid3R: Wide Field-of-View 3D Reconstruction via Camera Model Conditioning

Wid3R replaces the point-map output of multi-view feed-forward reconstruction with a decoupled "spherical-harmonics ray direction + radial distance" representation and conditions a fully shared backbone on a learnable camera model token, so a single network predicts dense point maps and camera poses from distorted fisheye and 360° images without calibration or undistortion, raising AUC@30° from π3's 40.94 to 74.61 on Zip-NeRF (fisheye) and from a previous best of 2.60 to 79.93 on Stanford2D3D (360°).

ZeroSplat: Generalized Referring Segmentation in 3D Gaussian Splatting

ZeroSplat converts category and localization priors from a vision-language model into Gaussian-level binary labels through multi-view masks and geometric constraints, requiring no per-scene semantic training or additional semantic feature storage and achieving 50.8 and 41.2 mIoU on GR-LERF and GR-ScanNet.

3D Scene-Adaptive Trajectory-Controllable Human Image Animation with Camera Movement

3STC-HIA adjusts human altitude and orientation along a user-defined path in a unified 3D scene, then guides a pretrained video model with visible scene point clouds for training-free human and camera control, reducing Trajectory100 translation error from RealisMotion's 1.305 m to 0.478 m without a corresponding improvement in yaw accuracy.

4DGS360: 360° Gaussian Reconstruction of Dynamic Objects from a Single Video

4DGS360 uses confident 2D tracks as spatiotemporal anchors for 3D tracking, initializes motion with geometry that includes occluded regions, and then optimizes dynamic Gaussians without diffusion in its base pipeline, reducing mean LPIPS on five iPhone scenes from HiMoR's 0.4639 to 0.3877.

A Benchmark for Heterogeneous Stereo Deblurring with Physically- and Epipolar-constrained Cross Attention

The paper introduces HSD for asymmetric wide/ultra-wide smartphone deblurring and PECA for cross-view retrieval within a physically plausible, directional epipolar window, improving NAFNet from 32.16 to 32.92 dB on HSD without the cost of global matching.

A second-order theory of texture for depth from focus

The paper explains why apparently textureless surfaces can provide subjective speckle for depth from focus, and enhances that signal with narrowband filtering under ambient light, at the cost of longer exposure and with material- and illumination-dependent benefits.

AC3S: Adaptive Conditioning for 3D-Aware Synthetic Data Generation

This paper proposes AC3S, which dynamically attenuates the conditioning injection strength of ControlNet using an adaptive visual prompt modulator, supplemented by a multi-agent VLM system to generate text prompts consistent with geometric structures. While maintaining precise 3D pose alignment, this approach significantly enhances the realism and diversity of synthetic images, reducing the FID by 15.95 points compared to the baseline 3D-DST and achieving notable improvements in downstream classification and pose estimation tasks.

Agentic Collaborative Cognition for Zero-Shot 3D Understanding

This paper reformulates zero-shot 3D scene understanding as an iterative collaboration between a Planning Agent and a Perception Agent centered around an explicit holistic cognitive map. By actively planning key views, recording object attributes across viewpoints, and filtering candidates through feedback, the proposed method significantly outperforms existing zero-shot approaches on 6 benchmarks, including ScanRefer, Nr3D, SQA3D, and ScanQA.

AIMold: An Autonomous AI-based Pipeline for Complex Mold Design

AIMold combines orientation selection, coarse voxel assembly generation, latent geometry refinement, and CAD post-processing, raising auxiliary-component COV from the VecSet baseline's 21.24% to 57.29% on the MoldCAD test split while still requiring downstream manufacturing verification.

AirSplat: Alignment and Rating for Robust Feed-Forward 3D Gaussian Splatting

AirSplat fine-tunes only DA3's Gaussian prediction head, corrects misaligned supervision through training-time self-consistent pose alignment, and suppresses floaters with teacher-rated opacity constraints, raising PSNR from 20.78 to 23.08 dB on RE10K with 12 input views.

AnchorSplat: Fast and Structure Consistent Detail Synthesis for Gaussian Splatting

AnchorSplat treats a low-quality Gaussian asset as an attributed point cloud and directly synthesizes 3D detail through local point anchors and single-pass multi-Gaussian generation, reaching 36.57 dB PSNR on 3DGS-SR with a reported processing time of approximately 0.01 seconds per asset without accessing original multi-view images at inference.

AnyMatch: Supercharging Universal Multi-Modal Image Matching with Large-Scale Single-View Images

AnyMatch proposes a framework for synthesizing multimodal, multiview matching data from vast collections of single-view images at zero cost. Monocular depth estimation + 3D reprojection + diffusion inpainting enable geometrically consistent viewpoint transformations, while cross-modal image translation generates four modalities—infrared/depth/normals/events—and sample-level geometric consistency verification (SGCV) filters generation artifacts; after fine-tuning LoFTR/EDM/RoMa on the synthetic Any-syn dataset, cross-modal matching performance substantially surpasses existing synthesis methods such as MINIMA, with strong zero-shot generalization to unseen medical/remote-sensing modalities.

APT: Anchor-aligned Perturbations for Tamper Localization in Fully Regenerated Images

To tackle the vulnerability of proactive watermarks being erased during real-world fully regenerated (FR) diffusion inpainting, APT optimizes dense, vector-wise anchor-aligned perturbations in the VAE latent space, utilizing the intrinsic alignment disparity between synthesized foregrounds and reconstructed backgrounds to achieve accurate pixel-level tamper localization.

AquaStereo: Enabling Underwater Stereo Matching via Depth-Conditioned Diffusion and Geometry Self-Distillation

To tackle extreme data scarcity and severe feature degradation in underwater stereo matching, AquaStereo establishes a depth-conditioned diffusion pipeline with left-right consistency checking to synthesize geometry-faithful stereo datasets, combined with cross-domain geometry self-distillation from a frozen terrestrial teacher and a temporal perception-enhanced matcher, setting new zero-shot state-of-the-art accuracy without target-domain fine-tuning.

Argus: Metric Panoramic 3D Reconstruction for Indoor Scenes

Addressing reference drift and scale ambiguity in unordered panoramic image reconstruction, Argus introduces a learnable covisibility-guided reference selection module and overcomplete geometric factorization supervision, achieving state-of-the-art feed-forward metric indoor 3D reconstruction on the large-scale Realsee3D benchmark.

Articulated Object Reconstruction from Rest-State Observation

Rest2Art reconstructs multi-part articulated 3D objects directly from a single closed configuration by iteratively co-refining vision-language and segmentation predictions grounded on an explicit mesh, and estimating joint parameters from video diffusion motion hypotheses validated against geometric constraints.

Audio-Visual Camera Pose Estimation with Passive Scene Sounds and In-the-Wild Video

This paper proposes the first method that leverages passive scene audio to assist vision-based relative camera pose estimation. By performing late fusion of a spatial audio encoder (SAE)—composed of a direction-of-arrival (DOA) spectrum and a binauralized embedding—with the SOTA visual pose estimation model Reloc3r, this method achieves significant improvements in both real-world scenes and simulated datasets.

AutoCompass: Accurate Visual Localization on Public Maps by Learning from Weak Labels

AutoCompass introduces a weak-supervision framework for 2D neural map matching that eliminates heading label dependencies, absorbs large-scale GPS positional noise through chunk-based marginalization, and incorporates SfM relative pose constraints, achieving state-of-the-art visual localization on public maps without accurate absolute ground-truth labels.

Beyond Isolated Objects: Relationship-aware Open Vocabulary Scene Understanding via 3D Scene Graph Analysis

Addressing the issue that existing 3D open-vocabulary methods treat objects as isolated entities and struggle with ambiguous appearances, RelGraphOV extracts relationship-aware 3D scene graphs via multi-view VLM reasoning and applies an Adaptive Gated Dual-Stream Contextual GAT with hierarchical contrastive learning to decouple geometric and semantic propagation, substantially outperforming prior arts in fine-grained and long-tail 3D segmentation.

Broadband Wide Field of View Imaging with Computational Mirrors

Addressing severe chromatic aberrations in refractive lenses across the 400–1700nm VIS-SWIR spectrum alongside the debilitating Petzval field curvature in reflective mirrors, this work introduces Computational Mirrors—a hardware-software system pairing sparse focal stacks (2–4 exposures) with a physics-inspired affine-deformed PSF model (SeidelConv) and Plug-and-Play deconvolution to achieve full-FoV achromatic imaging on a single sensor.

CAM3R: Camera-Agnostic Model for 3D Reconstruction

Addressing the severe degradation of recent 3D foundation models on non-pinhole cameras due to rectilinear optics assumptions, CAM3R decouples pairwise geometry into spherical harmonic ray direction fields and radial distances with relative poses, and introduces ray-aware global alignment to achieve robust, calibration-free 3D reconstruction across panoramic, fisheye, and pinhole views.

Capacity-Controlled Multi-View Stylization of 3D Gaussian Splatting

This paper proposes a multi-view stylization framework for 3D Gaussian Splatting based on capacity-controlled optimal transport. By replacing greedy nearest-neighbor matching with semi-balanced optimal transport, it suppresses the "many-to-one" matching issue via column capacity constraints. Concurrently, it introduces a cross-view matching guidance mechanism to ensure style consistency between adjacent views, supported by geometric regularization to enhance reconstruction quality. The proposed method comprehensively outperforms existing approaches in style quality, content fidelity, and multi-view consistency.

CapFrame: Text-Instructed Viewpoint Grounding in 3D Gaussian Scenes via Geometric Pseudo Labels

Addressing the limitation that conventional language-guided 3D methods only determine what object to view rather than how to compose and orient it, CapFrame formalizes the Text-Instructed Viewpoint Grounding task and introduces a Retrieve–Translate–Refine framework that derives geometric pseudo labels from instructions and optimizes the 6-DoF camera pose via differentiable 3DGS rasterization.

CasaMaestro: Multi-View Panoramas for House-Scale 3D Reconstruction

CasaMaestro is the first feedforward model for house-scale, extrinsic-free 3D reconstruction from multi-view panoramas, using only 20 to 50 sparse indoor 360° captures to jointly predict metric depth and global camera poses for fast, metric point-cloud generation across entire homes.

Category-Level 3D Correspondence in Camera Space via Morphable Object Priors

Addressing the severe ambiguities of viewpoint, occlusion, and scale in single-view category-level 3D correspondence, this work introduces HouseCorr3D, the first large-scale camera-space benchmark with amodal and symmetry annotations, and Morpheus, a framework that disentangles canonical shape priors, instance affine deformations, and 6D pose to enable semantically aligned 3D correspondences without explicit correspondence supervision.

CoIn: Comprehensive 2D-3D Inpainting with Gaussian Splatting Guidance

CoIn establishes a bidirectional collaborative framework bridging 2D latent diffusion generation and 3D Gaussian Splatting (3DGS). By utilizing Reference-Adaptive GS (Ref-GS) to establish coarse geometry, Consistency Loss Guidance (CLG) via GS-based reference feature warping to enforce cross-view coherence during 2D diffusion denoising, and a patch-based Texture-Enhancing Discriminator (TE-D) to restore high-frequency fidelity, CoIn supports arbitrary-shaped masks (including loose bounding boxes) across both object removal and object insertion tasks.

CoMind: Understanding Collaborative Human Activity from Multiple Minds and Views

CoMind introduces an 81-hour multi-modal dataset capturing dual-participant egocentric and exocentric video, eye tracking, and dense 3D scene/object scans in unscripted kitchen collaboration, establishing three sequential Theory of Mind benchmarks: Joint Attention Estimation, Socially Conditioned Object Interaction Anticipation, and Collaborative Handover Prediction.

Complex-Valued 2D Gaussian Representation for Computer-Generated Holography

This paper introduces a compact hologram representation based on complex-valued 2D Gaussian primitives and a customized CUDA rasterization-propagation pipeline, reducing the parameter search space by 5:1 via Gabor's minimum space-frequency uncertainty while maintaining high-fidelity diffraction reconstructions across both smooth and random phase-only holograms.

Confidence-Based Mesh Extraction from 3D Gaussians

To prevent 3D Gaussian Splatting from faking view-dependent reflections with corrupted geometry, this paper introduces a self-supervised confidence framework that dynamically balances photometric and geometric losses, combines it with color and normal blending variance penalties, and decouples the D-SSIM luminance term, achieving state-of-the-art unbounded mesh extraction without heavy external priors.

COSY: Compositional 3DGS Synthesis for Disentangled Human Head Editing

COSY presents a compositional 3D Gaussian Splatting GAN architecture that decomposes human head generation into four independent sub-generators (hair, skin, glasses, and torso) guided by sparse color histograms and boolean flags without explicit geometric templates, enabling real-time, zero-leakage 3D attribute editing.

CSS-BA: Gate Guided Column Space Search for Bundle Adjustment

Addressing severe ill-conditioning and pose/calibration distortion in Schur Levenberg–Marquardt under low-parallax and near-rotational motions, CSS-BA preserves the original BA objective, residuals, and all parameters while restricting each LM update step to a gate-guided, Ritz-basis low-dimensional subspace, achieving robust geometric recovery without artificial priors.

DANTE-W: Diffuse Albedo Neural Texturing in the Wild

Addressing the baked-in shading and shadowing artifacts in classical multi-view mesh texturing of unstructured in-the-wild scenes, Dante-w lifts view-space diffusion priors onto a UV-space neural texture and leverages the frequency discrepancy between albedo and irradiance through physically principled neural rendering to recover high-fidelity, 3D-consistent diffuse albedo maps ready for standard graphics engines.

DAP: Doppler-aware Point Network for Heterogeneous mmWave Action Recognition

To tackle real-world distribution shifts in mmWave action recognition caused by diverse radar devices and frequency bands, this paper introduces UniMM-HAR, the first large-scale heterogeneous multi-source benchmark, alongside DAP-Net, which leverages Doppler priors for dual-space reparameterization and textual semantic alignment to achieve state-of-the-art cross-source generalization.

Denoising-GS: Gaussian Splatting with Spatial-aware Denoising

Denoising-GS reformulates 3D Gaussian Splatting optimization as a spatial-aware primitive denoising process, simultaneously refining mean positions and covariance structures while leveraging Fisher information uncertainty pruning to achieve SOTA novel view synthesis with 10%+ fewer primitives.

DETRAM: End-to-end DEtection, Tracking and Recovery of HumAn Meshes

DETRAM presents an end-to-end single-stage transformer framework that unifies multi-person detection, 3D SMPL mesh recovery, and temporal tracking within a single decoder pass using persistent detection, tracking, and user-prompt queries.

DF3DV-1K: A Large-Scale Dataset and Benchmark for Distractor-Free Novel View Synthesis

Addressing persistent floating artifacts and geometry corruption caused by transient occlusions in casual real-world captures, this paper introduces DF3DV-1K, the first large-scale benchmark of 1,048 indoor and outdoor scenes with paired clean ground-truth views, benchmarks 10 representative radiance field methods, and demonstrates DI2FIX, a plug-and-play 2D diffusion enhancer fine-tuned on the dataset that yields average gains of 0.96 dB in PSNR.

Differentiable Polarized Path Tracing

To resolve numerical breakdowns in path replay backpropagation caused by the rank-deficiency of polarimetric Mueller matrices, this paper presents an unbiased differentiable polarized path tracing framework using cached suffix radiance and hybrid recomputation, achieving constant memory usage and higher efficiency for material, normal, and optical element optimization.

DiffGI: Differentiable Geometry Images for High-Fidelity Thin-Shell 3D Generation

DiffGI integrates a continuous 2D Truncated Signed Distance Function (TSDF) into geometry images with an analytical Differentiable Marching Squares algorithm and geometry-aware normal rendering loss, achieving high-fidelity, subpixel-boundary thin-shell and non-manifold 3D surface generation within an ultra-compact \(32 \times 32 \times 4\) latent space.

Diffusion-Based Material Regularization for Physics-Based Inverse Rendering

This paper proposes regularizing physically disentangled inverse rendering with similarity kernels derived from diffusion-predicted G-buffers (albedo, roughness, metallic, and normals): rather than fitting these predictions as target values, it uses them to define material similarity between surface regions, applying joint bilateral filtering to encourage consistent optimized parameters within the same material region, thereby suppressing artifacts caused by lighting baked into materials and significantly outperforming SOTA methods such as Neural-PBIR and MaterialFusion on Stanford-ORB.

DiGS-Avatar: Single-Image Animatable 3D Human Reconstruction via UV-Space Diffusion

DiGS-Avatar reformulates single-image animatable 3D human reconstruction as an efficient 2D UV-latent completion task, supervising a single-view diffusion student via a multi-view teacher and injecting rich appearance cues through cross-attention to reconstruct a photorealistic, rig-ready 3D Gaussian avatar in only 0.71 seconds.

Director: Instance-aware Gaussian Splatting for Dynamic Scene Modeling and Understanding

Director tightly couples 4D dynamic Gaussian Splatting reconstruction with instance and linguistic semantic modeling at the primitive level, leveraging cross-view aligned SAM3 masks and MLLM temporal sentence embeddings to supervise learnable 8-dimensional semantic features on Gaussians, achieving photorealistic dynamic rendering, robust non-rigid tracking, and open-vocabulary 4D instance segmentation.

Discovering Geometric Biases in 3D Face Reconstruction: A Curvature-Aware Spectral Framework for Fairness Evaluation

Addressing the insensitivity of traditional Euclidean metrics to local high-frequency morphological discrepancies that conceal algorithmic bias, this paper introduces a Curvature Reconstruction Error (CRE) based on locally fitted Laplace-Beltrami operators alongside a spectral error clustering framework, uncovering and quantifying systematic age and ethnicity geometric biases in 3D Morphable Models and downstream reconstruction methods.

Do Flat Minima Improve Sparse Novel View Synthesis?

This paper fundamentally investigates the relationship between loss sharpness and generalization in sparse novel view synthesis, revealing that pursuit of uniform flat minima causes over-smoothing in fine details, and introduces Structure-Aware Sharpness Regularization (FASR) to adaptively balance detail preservation and floater suppression based on local image structure.

DR-GS: Physically-Based Deformable and Relightable 2D Gaussians

DR-GS establishes a unified Gaussian framework integrating physically-based inverse rendering, relighting, and deformation-aware manipulation, effectively resolving baked-illumination artifacts on glossy surfaces under dynamic motion through decoupled parameter learning and accelerated mesh ray tracing.

DreamEdit3D: Personalization of Multi-View Diffusion Models for 3D Editing

DreamEdit3D introduces a disentangled personalization representation learning framework for multi-view diffusion models, leveraging a two-phase optimization pipeline consisting of attention-aligned textual inversion and joint multi-view full UNet fine-tuning to achieve precise, identity-preserving object-level compositional 3D editing and high-fidelity mesh reconstruction.

DualDiff3D: Dual Structure-Appearance Diffusion Priors for Reliability-Enhanced 3D Gaussian Splatting

To eliminate blurring caused by conflicting structure and appearance when conditioning image diffusion on sparse reference views, DualDiff decouples them into dual diffusion branches coupled by Structure-Appearance Attention (SAA), while DualDiff3D embeds this into a reliability-enhanced Render-Refine-Optimize (RRO) loop to robustly enhance sparse-view 3DGS reconstruction.

Dynamic-Robust Photometric–Semantic Reconstruction for Open-Vocabulary 3D Scene Understanding

To tackle spatial feature misalignments caused by moving foregrounds in dynamic environments, SPAR introduces a Cross-View Dynamic Region Predictor alongside a joint semantic-geometric encoder-decoder that filters transient motion noise before latent aggregation and performs self-supervised, dynamic-region-aware reconstruction in a single forward pass.

Efficient and High-Quality Depth Estimation via Pixel-Space Diffusion with Linear Attention

Addressing the prohibitive quadratic complexity and multi-step sampling latency in high-resolution generative monocular depth estimation, Lapis presents the first linear-attention pixel-space diffusion framework with single-step x-prediction, bridging the fidelity gap via patch-level consistency and pixel-level refinement modules to deliver state-of-the-art accuracy, sharp boundary recovery, and up to 10.9x speedups.

Ego-Human Motion Prediction with 3D-Aware LLM

Ego3DLM casts egocentric motion forecasting into a 3D scene-aware language model that simultaneously decodes past and future 3D poses alongside natural language narrations in a single autoregressive pass, refined by multi-modal GRPO to ensure physical plausibility and semantic fidelity.

EgoGVAE: Ego-body Mesh Reconstruction via Guided Variational Autoencoder

Addressing the severe latency and computational cost of diffusion-based head-to-motion generation, EgoGVAE leverages a full-body motion variational autoencoder as a training-phase prior to guide the latent distribution of a head-conditioned network, achieving superior full-body mesh recovery with single-step sampling at over 50x faster inference speed.

Error-Driven Scene Editing for 3D Grounding in Large Language Models

To alleviate the severe visual and spatial biases in 3D-LLMs caused by language priors and dataset co-occurrence shortcuts, DEER-3D introduces an error-driven closed-loop framework that performs deterministic predicate-level failure attribution, synthesizes minimal counterfactual 3D scene edits, and constructs hierarchical question–answer pairs for iterative debiasing and robust geometric grounding.

Extending a Large View Synthesis Model for Multi-view Panoptic Segmentation

This paper presents a decoupled multi-view panoptic segmentation framework demonstrating that a frozen Large View Synthesis Model (LVSM) trained solely on RGB photometric reconstruction inherently learns input-agnostic geometric correspondence, enabling direct propagation of 3-bit binary-encoded panoptic labels to unobserved novel viewpoints without explicit 3D reconstruction or task-specific fine-tuning (reaching 33.56 dB PSNR and 0.5949 mIoU on ScanNet).

Fast and Compact 3D Gaussian Splatting with Polarized Opacity Prior

Addressing the severe model bloat and gradient leakage caused by the standard densify-then-prune paradigm in 3DGS, this paper proposes a synergistic optimization framework coupling a Polarized Opacity Prior (POP) with an L2 reconstruction loss, which drives informative Gaussians to full opacity and redundant ones to zero, slashing Gaussian counts by up to 90% and accelerating training by over 30% without sacrificing visual fidelity.

FeDepth: Federated Learning for Depth Estimation under Robot Heterogeneity

Addressing non-IID challenges caused by heterogeneous robot platforms, sensing environments, and depth ranges, FeDepth introduces a descriptor-based soft clustered federated learning framework that captures overlapping client distributions via symmetric Jeffreys divergence and redundant cluster merging, achieving superior generalization on unseen robot sequences.

FillGS: Filling Observation Gaps in 4D Gaussian Splatting via Viewpoint-Time Selection and Generative Refinement

To tackle motion blur and severe artifacts caused by spatiotemporal observation gaps in sparse-view 4D Gaussian Splatting, FillGS actively selects highly informative virtual spatiotemporal viewpoints using rendering sensitivity and motion-aware observation density, refines rendered views via a video diffusion model, and robustly fine-tunes the 4DGS field with matching-confidence and co-visibility masks.

FixAnything: 3D-Consistent Rendering Refinement via Video Generative Priors

FixAnything introduces a generalist rendering refinement framework that reformulates sparse-view artifacts across 3DGS, NeRF, meshes, and sparse point clouds as latent-space video-to-video translation, combining mask-aware conditioning with SfM-pose-driven Flow-DPO preference alignment to achieve photorealistic and 3D-consistent outputs via lightweight LoRA adaptation.

FlexiAvatar: Unified 3D Gaussian Human Avatars Under Arbitrary Body Visibility

Addressing the severe geometric distortion and texture drift caused by the full-body visibility assumption in existing monocular human avatar methods, FlexiAvatar introduces the first unified 3D Gaussian avatar reconstruction framework across arbitrary body visibility, combining rasterization visibility-aware pruning to block spurious gradient leakage with occlusion-robust SMPL-X tracking, part-specific residual refinement, and diffusion-based 360° rotation completion to outperform both dedicated and generalist baselines across full-body, upper-body, and head-only settings.

FoundDP: Revisiting Weak Disparity Observability in Dual-Pixel Depth Estimation

Addressing the structural collapse caused by weak disparity observability under dual-pixel (DP) narrow baselines in textureless and downsampled regions, FoundDP bridges physically grounded DP metric depth with global structural priors from a monocular foundation model (DAV2), leveraging defocus-aware feature alignment and adaptive spatial gating to deliver high-fidelity, metric-accurate dense depth estimation.

G2P: Gaussian-to-Point Attribute Alignment for Boundary-Aware 3D Segmentation

The paper proposes G2P, which establishes precise Gaussian-to-point correspondences using Mahalanobis distance and transfers opacity and scale attributes from 3D Gaussian Splatting to point clouds. This introduces appearance cues while preserving the original point geometry, effectively reducing segmentation confusion between objects with similar geometry but different appearances.

GARDEN: Gravity-Aligned Reconstruction of Disentangled ENvironments from RGB images

GARDEN presents an RGB-only physically grounded scene factorization framework that resolves gauge ambiguity via a learned Gravity-View coordinate system, reconstructs scene-specific rigid meshes with gravity-constrained 6-DoF poses, and eliminates duplicate geometry through conditional 3D point classification to deliver simulation-ready hybrid representations.

Generating Multi-view Adversarial Examples for Visual Geometry Grounded Transformer

Addressing the security vulnerabilities of the 3D vision foundation model VGGT in multi-view reconstruction, this paper introduces MVAP-G, the first feed-forward multi-view adversarial perturbation generator that produces visually imperceptible, cross-view consistent perturbations in a single forward pass without test-time iterative optimization, collapsing reconstructed 3D point clouds and disrupting downstream geometric poses in milliseconds.

Generative Relightable Avatars

Generative Relightable Avatars (GRA) decomposes full-body human relighting into explicit physics-grounded coarse rendering and flow-matching-based video generative refinement, achieving photorealistic, temporally coherent free-view avatars with rich dynamic wrinkle patterns under arbitrary novel illuminations.

GenHOI: Generalized Hand-Object Pose Estimation with Occlusion Awareness

To overcome severe occlusion and shape variation in monocular 3D hand-object pose estimation, GenHOI integrates InstructBLIP-derived hierarchical textual semantic prompts, multi-modal masked reconstruction across RGB-point cloud-text, and hand structural/SDF priors to achieve 89.34% ADD-0.5D accuracy on unseen objects in the DexYCB S3 benchmark.

GenSP: Consistent Spherical Parameterization via Learning Shape Generative Models

GenSP proposes a data-driven method to learn a neural generative model that continuously deforms a unit sphere into arbitrary genus-0 shapes, thereby obtaining cross-shape consistent and low-distortion spherical parameterization via inverse mapping. On ShapeNet, it improves isometric distortion by over 14× and consistency by over 3.6× compared to existing methods.

Geo-ID: Test-Time Geometric Consensus for Cross-View Consistent Intrinsics

Geo-ID proposes a training-free, test-time framework that couples the predictions of independently run pre-trained single-view intrinsic diffusion models using sparse 3D geometric correspondences. It aggregates these predictions into a confidence-weighted, voxel-level robust consensus, which is back-projected and injected as a guide into the subsequent diffusion sampling process via gradient updates on latent variables. This enforces consistency in albedo, roughness, and metallicity across views without degrading single-view decomposition quality, and enables direct use in downstream editable neural scene representations.

Geometric Foundation Model Distillation for Efficient Lunar 3D Reconstruction

Addressing the stringent onboard compute limitations in planetary exploration, this paper presents a compact knowledge distillation framework for large 3D foundation models (MASt3R) on uncalibrated lunar stereo reconstruction, leveraging truncated SVD decoder weight initialization and margin-filtered feature alignment to compress the model by 4.4× to 7.3× with a >2× speedup while incurring only a ~15% Chamfer error increase and outperforming sparse ground-truth supervision.

Geometrically Consistent Multi-View Scene Generation from Freehand Sketches

Addressing the severe geometric impoverishment and non-linear spatial distortions of freehand sketches, this paper presents the first single-stage framework for sketch-to-multi-view scene generation, coupling lightweight parallel camera adapters with sparse SfM correspondence supervision to synthesize geometrically consistent 360° novel views without reference photographs or per-scene optimization.

Geometry-Aware Style Transfer in 3D Gaussian Splatting

This paper proposes a geometry-aware 3DGS style transfer framework. By alternately updating color and geometry parameters via decoupled optimization and introducing a multimodal (RGB + depth + edge) contrastive feature matching loss (GCFM) to guide the transfer of style structure and geometric texture, it significantly outperforms existing 3DGS style transfer methods in terms of style fidelity and multi-view consistency.

Global Pose Control for Generative View Synthesis in Normalized Object Coordinate Space

Reformulates novel view synthesis as instruction-based image editing in Normalized Object Coordinate Space (NOCS), conditioning on absolute Plücker ray maps via regional attention and text-defined canonical frames to achieve high-fidelity, unposed-input global camera viewpoint control.

Graph-GSReg: Leveraging 3D Scene Graphs for Gaussian Splatting Registration

Lifts unstructured 3D Gaussian Splatting scenes into object-level 3D scene graphs, solves robust rigid alignment via TRIMs distance consistency and maximum clique search without pretraining, and seamlessly eliminates occlusion artifacts and floaters through self-supervised test-time optimization.

Gravity-aware partially calibrated absolute pose estimation from affine- or rotation-covariant features

Addressing absolute pose estimation with unknown focal length (PnPf), this paper incorporates IMU gravity priors and local geometric attributes of feature descriptors (affine or rotation-covariant) to derive novel polynomial constraints and construct two efficient solvers—UP1PfAC from a single affine correspondence and UP2PfORI from two orientation-covariant features—slashing sample requirements to 1 or 2 features while drastically accelerating RANSAC robust localization and focal length recovery.

GridFlow: Structured Latent Flow for Seamless City-Scale 3D Point Cloud Generation

GridFlow introduces a multi-stage framework combining a Grid-Aligned VAE with conditional rectified flow to generate dense (\(10^5\) points per tile), seamless, colored 3D point clouds at city scale conditioned on satellite imagery and auxiliary remote sensing data.

GrowFields: Compositional 4D Neural Fields for Topology-Changing Plant Growth

GrowFields decomposes longitudinal plant point cloud series into canonical organ coordinate frames and learns a shared continuous 4D SIREN velocity field conditioned on per-organ latent codes, achieving high-fidelity 4D reconstruction and leaf-tip tracking under topology-changing and asynchronous plant growth.

HAT-4D: Lifting Monocular Video for 4D Multi-Object Interactions via Human-Agent Collaboration

HAT-4D proposes the first agentic framework for monocular-video 4D multi-object interaction reconstruction. It uses an interaction knowledge graph (IKG) as a causal engine to drive 3D generation, spatial composition, and memory-augmented 4D propagation, while introducing multi-level human-in-the-loop (HITL) collaboration to address depth ambiguity and occlusion. On the MVOIK-4D benchmark, it comprehensively outperforms existing monocular 4D baselines, and a small amount of human feedback substantially improves interaction reconstruction quality.

Head Avatars with Dynamic Explicit Hair

DynHair presents an animatable head avatar combining explicit strand geometry with 3D Gaussian Splatting, using a temporal LSTM-FiLM network conditioned on head angular velocity, acceleration, and relative gravity alongside physical elastic regularization to generate realistic, temporally coherent hair dynamics from multi-view videos.

Hierarchical 3D Scene Graph Construction and Belief-based Planning for Semantic Navigation

This paper proposes a zero-shot semantic navigation framework that online incrementally constructs an object-zone-region hierarchical 3D scene graph (HSG) as structured global memory. It fuses LLM semantic priors with exploration evidence to maintain hierarchical belief states, and uses POUCT within an HSG simulator for finite-step forward rollouts to explicitly evaluate the long-horizon expected returns of macro-actions, improving long-range navigation SR/SPL by 9.4%/5.0%, respectively.

High-Resolution Artwork Outpainting with Global Blueprint Guidance and Layout Control

This paper presents BlueOut, a two-stage diffusion framework for high-resolution artwork outpainting that plans global composition with bounding-box layout control in Stage 1 and synthesizes high-resolution local patches in parallel in Stage 2 using forward diffusion low-frequency initialization, achieving superior visual fidelity, explicit spatial control, and a 2.4× speedup.

HitMem: Hierarchical Temporal 3D Memory with Multi-Modal Context-Aware Retrieval for Dynamic Environments

Addressing unexpected object displacements caused by external human activities in dynamic environments, HitMem establishes a hierarchical 3D memory graph coupling carrier-carried topological relations with role-aware temporal decay, and initiates a two-stage context-aware retrieval integrating external agent trajectories and semantic class affinity to efficiently relocate displaced targets without blind global exploration.

HybridSim: A Physics–Learning Hybrid Digital Twin for mmWave Human Sensing

Addressing the exponential ray-tracing overhead and neural-implicit motion entanglement in radar simulation, HybridSim decouples electromagnetic wave propagation into an inverse-rendering direct path with microfacet BRDF and a 3D Gaussian Splatting indirect path with proxy virtual receivers, synthesizing high-fidelity intermediate-frequency mmWave signals at under 1 second per frame.

Hyper-Network Neural Functional Maps for Unsupervised Robust 3D Shape Matching

Addressing the failure of standard linear functional maps to align truncated spectral bases under topological noise and severe partiality, this paper introduces the first unsupervised neural functional map framework, where a Vision Transformer hyper-network dynamically predicts non-linear residual MLP weights conditioned on an initial linear functional map and is trained via an unsupervised spectral alignment loss.

InceptionGS: Generative Bootstrapping for Large-Scale Gaussian Splatting under Unstructured View Sampling

To tackle unstructured view sampling and severe view scarcity in large-scale scene captures, InceptionGS adapts a latent diffusion model on dynamically rasterized G-buffer geometry from an initial Gaussian splatting, and refines the 3DGS field via photometric blending and Langevin importance sampling, reducing average FID by 32% without requiring additional real-world captures.

Inclusive Interactive Collisions for Multi-View Consistent Compositional 3D Generation

I2C-3D presents a coarse-to-fine optimization framework for compositional 3D asset generation, using an Inclusive Interactive Collisions strategy to naturally gather Gaussian primitives into plausible contact zones and Multi-View Adaptive Score Distillation Sampling to resolve cross-view inconsistency and the Janus problem.

Incremental Online Scene Reconstruction by 3D Gaussian Triangulation

This paper introduces an online incremental scene reconstruction framework that unifies high-fidelity explicit mesh generation and photorealistic rendering by directly triangulating planar 3D Gaussian surfels guided by plane-based pulling constraints and dynamic historical freezing.

InFlux++: Real and Synthetic Data for Estimating Dynamic Camera Intrinsics

InFlux++ introduces a dual-track data suite comprising 1,841 procedurally generated synthetic videos (441K+ frames with thin lens breathing and realistic optical aberrations) and 334 high-resolution real-world videos (514K+ frames with rich translation and reverse large-FSF calibration), where finetuning SOTA dynamic calibration models on synthetic data significantly boosts focal length estimation accuracy.

InSpace: Structure-Aware 3D Indoor Scene Generation from a Single 360° Image

Addressing the limited field of view and lack of structural anchors in existing single-image 3D generation, InSpace proposes a three-stage flow matching framework that generates complete 3D indoor scenes—including architectural layouts (floors/walls) and disentangled textured assets—from a single 360° equirectangular projection (ERP) image using geometric initialization, view-selective cross-attention, and OBB-guided global-local hybrid attention.

Instant Expressive Gaussian Head Avatars at Over 100 FPS

This paper tackles the portrait animation trilemma among inference speed, 3D multi-view consistency, and facial expressiveness by introducing a feedforward Gaussian framework that deforms primitives in a high-dimensional feature space using decoupled motion basis vectors, achieving detailed expression transfer and strict 3D consistency at 107.31 FPS.

InterPet4D: A Multimodal 4D Human-Pet Interaction Dataset for Pet Motion Generation

Presents InterPet4D, the first large-scale multimodal 4D human-pet interaction dataset (comprising 6.83 million synchronized frames across 23 participants and 13 dogs of 11 breeds), and proposes InterPetMoGen, an autoregressive generation framework integrating bone-constrained PetVAE tokenization with Modality-Aware Attention.

JanusMesh: Fast and Zero-Shot 3D Visual Illusion Generation via Cross-Space Denoising

JanusMesh proposes a zero-shot, two-stage framework that leverages dual-branch denoising and SDF fusion in the TRELLIS structured 3D latent space to generate multi-semantic geometries, followed by view-constrained texture synthesis to assign corresponding appearances to different viewpoints, producing high-quality 3D visual illusion meshes in just 3 to 5 minutes.

KISS-GS: 3D Gaussian Splatting Compression Kept Simple

KISS-GS proposes a modular, decoder-first 3DGS compression framework that strictly decouples reconstruction, compaction, encoding, and adaptation, enabling 85× to 319× file-size reductions over vanilla 3DGS at matching visual quality while requiring only ubiquitous, browser-native 2D image decoders.

Large-Scale High-Quality 3D Gaussian Head Reconstruction from Multi-View Captures

HeadsUp proposes a scalable feed-forward method that compresses multi-view images from dynamic camera rigs into compact 2D latent variables via a cross-attention Transformer. These latents are then decoded into UV-parameterized 3D Gaussians anchored to a neutral head template. This decouples the output Gaussian count from the input image resolution and view count, achieving SOTA reconstruction quality on an internal dataset of 10,000+ subjects without requiring test-time optimization.

Last-Layer-Centric Feature Recombination: Unleashing 3D Geometric Knowledge in DINOv3 for Monocular Depth Estimation

Addressing the flaw in conventional monocular depth estimation pipelines that uniformly sample ViT intermediate layers and ignore geometric heterogeneity, this paper reveals that deeper layers in DINOv3 possess stronger 3D predictive capacity, and proposes a Last-Layer-Centric Feature Recombination (LFR) module that adaptively fuses complementary intermediate layers based on minimal-similarity selection to achieve state-of-the-art results across standard and zero-shot benchmarks.

Learning Global Camera Poses from Noisy View-Graphs for Structure from Motion

VGPA (View-Graph Pose Averaging) introduces a permutation-equivariant, edge-conditioned GNN that directly aggregates noisy pairwise relative poses across view-graphs to regress global camera extrinsics; trained via an unsupervised relative-pose consistency objective and combined with decoupled triangulation and robust bundle adjustment, it scales to collections of over 1,000 images with high registration coverage, matching classical SfM accuracy while delivering substantial speedups.

Learning Implicit Constitutive Laws for Dynamic 3D Gaussian Splatting from Monocular Videos

To tackle severe geometric and material ambiguities when inferring physics from monocular video, GCA unifies LoRA-based adaptation with scale-invariant rank-based depth geometric anchors and soft constitutive prior regularization, achieving robust implicit constitutive modeling for dynamic 3D Gaussians.

Learning Manifolds in High-D Point Embedding for Anisotropic Surface Approximation from Unstructured Point Clouds

To tackle the challenge of estimating anisotropic metric fields on unstructured point clouds without connectivity, HD-PEA introduces a barycentric-invariant isometric transfer scheme to train Point Transformer V3 for 8D manifold embedding, followed by tangent subspace particle relaxation and high-dimensional restricted Voronoi cell (RVC) clipping, directly reconstructing compact, curvature-aligned anisotropic meshes from raw point clouds.

Learning Spectral and Polarimetric Clues for One-to-Multimodal Novel View Synthesis

Addressing the prohibitive acquisition costs of specialized sensors, SPoILeR decouples implicit radiance fields into shared bases and scene-specific coefficients with multi-scene pre-training, enabling photorealistic and strictly multi-view consistent synthesis of multispectral, infrared, and polarimetric novel views using only RGB supervision.

Learning Video Dynamics with Predictive Differentiable Rendering

This paper proposes the Predictive Differentiable Rendering (PDR) framework, which appends a lightweight PredGS adapter to existing pixel-space predictors to map coarse predictions to 2D Gaussian parameters. These parameters are rendered into continuous-space predictions via a CUDA-accelerated differentiable renderer and then fused with pixel predictions, significantly alleviating the over-smoothing issue of deterministic video prediction with minimal computational overhead.

LEGO: Leveled Language Gaussian Splatting

Addressing the viewpoint dependency and semantic-scale decoupling in existing 3D language Gaussian splatting, LEGO self-adaptively re-grades multi-view SAM granularities into unified 3D-consistent structural levels, performs decoupled feature distillation, and constructs level-wise language scene graphs to empower LLMs for complex spatial reasoning.

LiteMatch: Lightweight Zero-Shot Stereo Matching via Cost Volume Stabilization

Addressing the severe computational burden and cross-domain vulnerability of deep stereo models caused by heavy 3D cost volume regularization or bulky monocular depth foundation priors, LiteMatch stabilizes matching features and cost volumes at the source using a dual cross-view/high-frequency encoder and a voxel-wise binary cross-entropy CVC-Loss, delivering state-of-the-art zero-shot generalization and real-time inference with only 3.36M to 9.58M parameters.

LivingWorld: Interactive 4D World Generation with Environmental Dynamics

To tackle the static nature of existing interactive 3D world outpainting, LivingWorld models continuous environmental dynamics via geometry-aware Kabsch alignment and a compact hash-based 3D motion field, achieving globally coherent 4D world generation with bidirectional motion propagation in just 12 seconds per step.

LSRM: High-Fidelity Object-Centric Reconstruction via Scaled Context Windows

Addressing the persistent fidelity gap where feed-forward 3D reconstruction falls behind dense-view optimization in fine-grained appearance, LSRM scales object and image context windows by 20× and >2× using native sparse attention, geometric 3D-aware routing, and block-aware sequence parallelism to achieve state-of-the-art reconstruction and inverse rendering.

LUNA: Learning Universal 3D Human Animation Beyond Skinning

LUNA proposes the first end-to-end 3D human animation framework that completely abandons Linear Blend Skinning (LBS). It directly maps 2D driving signals (RGB images, keypoints, hand-drawn sketches, or even unseen character images) to 3D Gaussian deformations via a decoupled Transformer animator. During training, LBS distillation is utilized to provide structural priors to prevent geometric collapse. During inference, it completely bypasses LBS constraints, achieving zero-shot cross-identity and cross-modal 3D human animation.

Map-Det3D: Metric Feed-Forward 3D Reconstruction Prior for Multi-view 3D Object Detection from Streaming Inputs

Map-Det3D repurposes a feed-forward metric 3D reconstruction model (MapAnything) as the geometric encoder within a detection transformer, directly decoding up-to-scale 3D boxes from streaming RGB sliding windows and recovering absolute metric dimensions via an explicitly predicted scale factor, which bypasses error-prone 2D-to-3D lifting and achieves state-of-the-art results on CA-1M and zero-shot ScanNet.

MeGAS: Thermomechanical Dynamic Gaussian Splatting for Thermophysical Scene Editing

MeGAS integrates thermodynamic phase-change dynamics (heat conduction, convection, melting/solidification) into the 3D Gaussian Splatting framework, enabling the reconstructed real-world scenes to undergo controllable, physics-based thermomechanical deformation, while maintaining rasterization quality under extreme large deformations through a topology-adaptive rendering strategy.

Memory Tree Guided Key Frame Querying for Efficient 3D Question Answering

To tackle the computational explosion and severe latency of exhaustive per-query visual search in long-form embodied 3D question answering, this paper presents MemTree3D—a hierarchical scene memory tree constructed in real time (25+ FPS) from 6-DoF camera poses and object tracking—allowing an LLM to generate spatial-temporal cues to retrieve top-3 key frames, outperforming uniform sampling GPT-4o on OpenEQA by 17.4% with a 69.2% retrieval speedup.

MemPose: Category-level Object Pose Estimation with Memory

Addressing the limitation of static shape priors and fixed network parameters in adapting to diverse object instances, MemPose incorporates an external, non-parametric category-level geometric memory buffer into the pose estimation pipeline, achieving state-of-the-art 9-DoF accuracy with near-zero learnable parameter overhead.

MessyKitchens: Contact-rich object-level 3D scene reconstruction

Addressing the lack of physical plausibility and severe inter-object penetration in single-view object-level 3D reconstruction, this paper introduces the MessyKitchens benchmark with millimeter-level registration accuracy alongside a Multi-Object Decoder (MOD) that refines object poses via joint cross-attention, significantly enhancing contact fidelity and scene-level geometry.

MetaView: Monocular Novel View Synthesis with Scale-Aware Implicit Geometry Priors

MetaView establishes a monocular novel view synthesis framework under large viewpoint changes without explicit 3D reconstruction, by injecting intermediate feed-forward geometry priors into a frozen MM-DiT via parallel attention and embedding metric depth with camera poses into scale-aware RoPE, effectively resolving both scale drifting and reconstruction degradation.

Mitigating Positional Leakage in 3D Masked Autoencoders for Robust Representation Learning

MPL-MAE resolves the severe positional leakage issue in 3D masked autoencoders—where decoders over-rely on continuous metric coordinates instead of encoder semantics—by introducing order-isomorphic recalibrated positional embeddings, a Gumbel-Softmax gated positional interface, and leakage regularization.

MMDiff: Extending Diffusion Transformers for Multi-Modal Generation

Addressing the issue where diffusion transformers discard rich perceptual representations during rendering, MMDiff transforms a frozen DiT into a multi-modal generator using adaptive multi-timestep feature fusion, concept-driven attention, and complementary discriminative features to accurately predict semantic segmentation, saliency, and depth alongside generated images with lightweight decoder heads.

MoCA3D: Monocular 3D Bounding Box Prediction in the Image Plane

MoCA3D bypasses the reliance on camera intrinsics in conventional monocular 3D lifting by reformulating 3D bounding box prediction as an image-plane dense prediction task, achieving superior corner reprojection accuracy and competitive 3D performance with 57x fewer trainable parameters (19.0M).

Monocular Models are Strong Learners for Multi-View Human Mesh Recovery

MonoMV presents a training-free multi-view human mesh recovery framework that leverages strong priors from pretrained single-view HMR models, initializing a robust "Virtual View" and refining it via test-time adaptation guided by anatomical landmarks, cross-view consistency, and monocular prior regularization to achieve SOTA accuracy under both calibrated and calibration-free setups.

MSVS-VAE: Multi-Scale Anchored VecSet for High-Fidelity 3D Reconstruction

MSVS-VAE tackles the pervasive fidelity and efficiency bottlenecks of set-based 3D representations by hierarchically densifying anchored VecSet latents via point-shuffle upsampling and substituting global cross-attention with localized AVS-Conv, delivering voxel-level reconstruction fidelity with \(\sim\)10\(\times\) higher compactness and up to 14.4\(\times\) faster decoding.

Multi-THuMBS: Multi-person Tracking of 3D Human Meshes Beyond Video Shots

Addressing identity drift and trajectory fractures across abrupt camera shot transitions in real-world videos, Multi-THuMBS constructs a shared 3D scene anchor at shot boundaries using a geometric foundation model, linking multi-person trajectories via hybrid 3D-spatial Re-ID and global spatiotemporal smoothing.

Multi4D: High-Fidelity Dynamic Gaussian Splatting via Multi-Level Competitive Allocation

Multi4D explicitly decomposes a dynamic scene into three subsets of Gaussians: "static structure + persistent dynamic geometry + transient appearance", allowing them to competitively explain photometric residuals under shared rasterization. This concurrently preserves long-term motion consistency and high-frequency appearance details. Relying on 25\(\times\) fewer dynamic Gaussians than 4DGS, it achieves SOTA rendering quality and real-time frame rates, while casually enabling 4D segmentation with a 10\(\times\) speedup.

MV-GEL: Language-Driven Multi-View Geometric Entity Localization on Meshes

Addressing severe viewpoint occlusions and topological boundary ambiguity in 3D CAD grounding, MV-GEL decouples prompt-conditioned geometric view ranking (GELviews) from domain-adapted 2D reasoning segmentation (LISA-CAD), accurately localizing natural-language-queried faces and edges directly on unstructured polygon meshes via geometry-aware ray casting.

MVFusion-GS: Motion-Variance Guided Temporal Attention for High-Quality Dynamic Gaussian Splatting

To resolve background contamination and inaccurate foreground deformation caused by the lack of explicit motion awareness in dynamic 3D Gaussian Splatting, this paper introduces a plug-in refinement framework combining global motion-variance trajectory priors with local temporal cross-attention.

NeuroRefiner: Morphology-Aware Multi-Agent Refinement for 3D Fluorescence Microscopy Neuron Segmentation

Addressing extreme signal sparsity, fragile filamentous morphology, and irreversible errors in single-pass 3D fluorescence microscopy neuron segmentation, NeuroRefiner presents a closed-loop multi-agent architecture where Global Inspector, Refinement Advisor, and Change Validator collaborate to guide a dedicated 3D cross-modal editing network (TopoRefineNet) for iterative, voxel-level topological error correction.

NoPA: Non-Parametric Online 3D Scene Graph Generation

Addressing geometric detail loss and severe under-merging caused by single-Gaussian parametric assumptions in online 3D scene graph generation, NoPA introduces a fixed-size non-parametric particle representation with a two-stage "Hellinger pre-filter + MMD refinement" online merging strategy and affinity-based relationship propagation, doubling relationship recall while preserving real-time frame rates.

Objects as Audio-Visual Modal Sound Fields

Audio-Visual Modal Sound Field (AV-MSF) couples 3D Gaussian Splatting visual geometry priors with compact physical modal parameters to reconstruct continuous object-level impact sound fields from multi-view RGB images and few-shot recordings, enabling novel-position sound synthesis, contact localization, and text-guided sound editing.

OmniDS: Dual-Stream Context Fusion for Omnidirectional Depth from Fisheye Cameras

Addressing visibility conflicts and rigid projection ambiguities in wide-baseline multi-fisheye setups, OmniDS combines a dual-stream CNN and frozen DINOv3 dynamic ERP context fusion with a multi-view consensus volume, and achieves real-time ~10 FPS inference via lightweight MobileNet distillation.

OmniPoint: Universal Monocular Metric Pointcloud from Any Camera

OmniPoint introduces the first unified monocular metric point cloud estimation framework that handles arbitrary camera geometries (pinhole, fisheye, 360° panoramic) by decoupling ray direction from radial distance, incorporating bidirectional 3D data augmentation, and dynamically injecting geometric priors via input-state embeddings and Gaussian smoothing.

One Scene, Two Depths: Probing Geometric Ambiguity in Monocular Foundation Models

Addressing the single-scalar collapse in monocular depth foundation models under layered scenes, this paper introduces the sparse two-layer ordinal benchmark MD-3k to diagnose depth-layer preferences and demonstrates that training-free Laplacian Visual Prompting (LVP) can modulate frozen models to emit complementary geometric hypotheses.

OneWorld: Taming Scene Generation with 3D Unified Representation Autoencoder

OneWorld discards conventional 2D/video latent generation paradigms by constructing a 3D Unified Representation Autoencoder (3D-URAE) that injects appearance and distills semantics, enabling direct diffusion within a geometry-aware 3D latent space while enforcing cross-view correspondence and mitigating exposure bias via manifold-drift forcing.

Online Segment 3D Gaussians via Launching Virtual Drones

SAGO reformulates interactive 3D Gaussian Splatting segmentation as an online Next-Best-View (NBV) planning task driven by virtual drones within a Markov process, completely eliminating the time-consuming per-scene setup stage and enabling clean 3D asset extraction with sub-second latency (0.4–0.9s).

OP3DSG: Open-Vocabulary Part-Aware 3D Scene Graph Generation for Real-World Environments

OP3DSG proposes a unified 3D scene graph generation framework. By leveraging knowledge-guided part detection, part-aware multi-view fusion, and a geometrically initialized prior graph combined with verification-gated LLM reasoning, it jointly models objects, interactive parts, spatial relations, functional relations, and affordances. On the self-constructed UniGraph3D benchmark, its part-node recall outperforms prior methods by an absolute margin of over 31 points, and it can directly serve as a perception backend for robotic QA, navigation, and task planning.

Open-Vocabulary 3D Object Detection with Co-Distillation Discovery and Dual Guidance Robust Training

To tackle inaccurate localization and mismatched semantics in novel object discovery alongside indiscriminate supervision during detector training, Co-3DGT unifies 2D-3D co-distillation bipartite matching with scene-aware uncertainty regularization and LLM-guided hierarchical alignment, substantially advancing unseen category detection on SUN RGB-D and ScanNetV2.

OpenGround: Planning-based Online Perception for Open-World 3D Visual Grounding

Addressing the failure of existing zero-shot 3D visual grounding methods on undefined targets outside the predefined Object Lookup Table (OLT), OpenGround decomposes complex queries into an ordered task chain of sub-goals and leverages grounded context anchors to perceive novel objects online, achieving superior grounding performance on the newly created OpenTarget benchmark.

OVGGT: O(1) Constant-Cost Streaming Visual Geometry Transformer

OVGGT introduces a training-free streaming visual geometry transformer framework that bounds both memory footprint and per-step computation to an \(O(1)\) constant budget via FFN-residual-based Self-Selective Caching with spatial smoothing and Dynamic Anchor Protection, eliminating out-of-memory failures while outperforming full-cache baselines.

P-CORE: Self-Supervised Surface Consistency for Point-Based Neural Editing

Addressing the severe holes and surface tearing that plague point-based neural renderers under non-rigid deformations, P-CORE introduces a self-supervised surface consistency framework requiring no dynamic ground-truth data or geometry proxies, leveraging operator commutativity between deformation and surface prediction alongside dual ray deformation modes to deliver continuous, hole-free zero-shot editing with only 30K extent-free points.

Pano3D: Unified 3D Reconstruction and Panoptic Segmentation

Pano3D directly appends a Mask2Former-style set-based mask decoder to a Feedforward Reconstruction Model (FRM, e.g., MUSt3R / Pi3), and jointly fine-tunes the geometric decoder using both geometric and semantic losses. This represents the first approach to simultaneously output dense point clouds and 3D panoptic segmentation in a single feedforward pass without requiring external 2D models or post-processing clustering. It significantly outperforms state-of-the-art (SOTA) methods in semantic segmentation mIoU on ScanNet (+16.6), ScanNet200, and ScanNet++.

PanoLess: Environment Reconstruction from Partial Reflective Views

PanoLess reconstructs high-frequency distant environment illumination from partial reflective views on a single side of shiny objects via surface-aligned 2D Gaussian surfels, unattenuated neural cubemap deferred shading, and an online visibility confidence map, improving environment map PSNR by nearly 5 dB over prior reflection-aware baselines.

PASTEL: Panoramic Alignment for Monocular 4D Scene Reconstruction

PASTEL projects monocular video into a 2D spherical panoramic space to transform unconstrained 6-DoF trajectory exploration into a structured 2D directional search, combining static-dynamic point aggregation with strategic confidence masking to reliably extrapolate unobservable 4D dynamic scenes beyond camera boundaries.

Physically Grounded 3D Generative Reconstruction under Hand Occlusion using Proprioception and Multi-Contact Touch

This work integrates monocular RGB vision, proprioception (hand pose), and multi-contact touch within a flow matching diffusion framework to generate the metric-scale SDF of an object under severe hand occlusion. It guides shape reconstruction using two physical losses—non-penetration and contact consistency—during both training and inference sampling stages, yielding physically feasible outputs and implicitly estimating the object pose.

Physically Grounded Monocular Depth via Nanophotonic Wavefront Encoding

By engineering a birefringent metalens to passively encode metric depth into depth-dependent rotational shifts across orthogonal polarization channels in a single monocular capture, combined with a zero-overhead pseudo-RGB input adaptation strategy, this paper directly fine-tunes depth foundation models (Depth Anything V2) to deliver highly accurate metric monocular depth without active LiDAR sensors.

PhysMani: Physics-principled 3D World Model for Dynamic Object Manipulation

Addressing the failure of existing VLAs and video world models to manipulate fast-moving targets due to the absence of explicit 3D geometry and physical laws, PhysMani couples an online-optimized divergence-free 3D Gaussian velocity field with a future-aware action policy, achieving state-of-the-art dynamic manipulation success rates across 16 benchmark tasks and real-world robot experiments.

PIC: Revisiting INR for Image Coding with Fast Encoding and Sub-Millisecond Decoding

Addressing the severe encoding bottleneck of implicit neural representation (INR) image codecs caused by per-image overfitting, PIC introduces a feedforward hypernetwork modulation architecture with a hardware-aligned ultra-lightweight decoder that computes multi-scale latents and network weights in a single forward pass, achieving 20 FPS practical encoding, 2000 FPS (0.42 ms) sub-millisecond decoding, and rate-distortion performance superior to JPEG.

PixVOD: Pixel-Distributed Direct Visual Odometry and Depth Estimation

Addressing the on-sensor computing paradigm (Pixel Processor Array), PixVOD introduces the first fully pixel-distributed direct visual odometry and depth estimation framework using Gaussian Belief Propagation (GBP) and keyframe anchoring, achieving global 6DoF consensus and metric depth recovery solely via local inter-pixel message passing.

Point Ladder Tuning: Parameter-Efficient Hierarchical Adaptation for 3D Point Cloud Understanding

Point Ladder Tuning (PLT) constructs a multi-resolution local geometric feature pyramid directly from raw point clouds and feeds instance-aware dynamic prompts back into a frozen 3D Transformer backbone via selective local-global fusion, outperforming full fine-tuning with merely 2.71% trainable parameters on classification and dense prediction.

PointSplat: Compact Gaussian Splatting via Human-Centric Prediction

PointSplat introduces a human-centric feed-forward 3D Gaussian Splatting framework that directly infers compact Gaussian primitives in 3D space from a space-carved, ray-cast geometric proxy, achieving superior novel view synthesis quality over view-centric baselines while utilizing only about 33% of the Gaussian primitives.

PRISM-VO: Scale-Aware Visual Odometry Using Photometric Plenoptic Bundle Adjustment

Tailored to the optical physics of focused plenoptic cameras, PRISM-VO formulates an end-to-end photometric plenoptic bundle adjustment framework that jointly optimizes single-frame micro-lens disparity depth priors and sliding-window multi-view temporal photometric consistency, enabling drift-resilient, metric-scale visual odometry without external sensors.

PRISM3D: Probabilistic Refinement and Robust Initialization for Physically Consistent Scene Modeling under Extreme Motion Blur

Addressing the complete breakdown of traditional SfM under severe motion blur, PRISM3D couples deep dense tracking (VGGSfM) for topological initialization, 3DGS-MCMC probabilistic sampling to populate sparse priors, and continuous \(SE(3)\) Bézier trajectory modeling for physically consistent joint optimization—achieving state-of-the-art blind 3D scene reconstruction without sharp-image supervision.

Pro-Pose: Unpaired Full-Body Portrait Synthesis via Canonical UV Maps

Pro-Pose develops an unpaired full-body portrait synthesis framework operating in canonical UV texture space with a donor-based reposing strategy that eliminates pose-texture boundary leakage, enabling photorealistic, pose-controllable avatar generation in standardized minimal garments from a single in-the-wild image and acting as an occlusion-free geometric canvas for virtual try-on.

Progressive Pose-Guided 4D Animal Reconstruction from Monocular Video

This paper proposes a progressive test-time optimization framework based on 3D Gaussian Splatting. By decoupling skeletal motion from non-rigid deformation, it reconstructs high-fidelity 4D animal models from monocular videos, requiring only coarse-grained shape priors to generalize to diverse species.

PUF: Plug-and-Play Uncertainty-Aware Fusion for Online 3D Scene Graph Generation

Addressing the failure of existing online 3D scene graph generation methods to account for observation truncation, 2D model ambiguity, and 3D spatial representation uncertainty under deterministic fusion pipelines, PUF introduces a training-free plug-and-play framework combining probabilistic node association, dual Dirichlet evidence accumulation, and class-conditional relation priors, boosting relationship recall on 3DSSG by 18.1 points while operating at a real-time latency of 15 ms per frame.

PWM-ArtGen: Part World Model for Articulated Object Generation

PWM-ArtGen proposes the first part world model for single-image articulated 3D object generation, coupling action-induced visual dynamics and kinematic parameters through decoupled diffusion timesteps, visual dynamics regularization, and co-training on 19.7k unannotated photorealistic images to achieve robust zero-shot articulated generation.

Racing in Volume with Flow Ensembles

Addressing catastrophic tracking breakdowns in streaming 4D Gaussian reconstruction under extreme outdoor vehicle displacements, FastFlowGS fuses multi-scale correspondence ensembles via geometric uncertainty and a Kalman temporal prior, achieving the first robust streaming reconstruction on the newly released Monaco4D benchmark.

RADmesh: Remesh-Aware Mesh Deformation

Addressing the severe triangle distortion and structural constraints in generative explicit mesh deformation under noisy visual supervision, RADmesh integrates a vertex-based 6D scale-rotation deformation quantity with coarse-to-fine periodic isotropic remeshing and barycentric optimizer state interpolation, enabling large-scale local growth and global detailing while strictly preserving isotropic element quality.

RayDer: Scalable Self-Supervised Novel View Synthesis from Real-World Video

RayDer unifies camera pose estimation, scene reconstruction, and novel view rendering into a single feed-forward transformer backbone while absorbing video dynamics as a nuisance factor via state dropout, breaking the static-scene data barrier and unlocking predictable power-law scaling on unconstrained real-world video.

REALM: An RGB and Event Aligned Latent Manifold for Cross-Modal Perception

REALM projects asynchronous event voxel grids directly into the latent manifold of a frozen RGB foundation model (DUNE) via a lightweight convolutional embedder and parameter-efficient LoRA adapters, enabling the direct, zero-shot application of frozen image-trained decoders (such as MASt3R) to raw event streams without any event-specific decoder training.

ReconSplat: Generalizable 3D Scene Reconstruction Beyond Observed Views

ReconSplat presents a feed-forward 3D scene reconstruction and view extrapolation framework that rasterizes 3D variational Gaussian distributions into 2D latent priors, guiding a multi-view latent diffusion model (MV-LDM) to jointly synthesize photorealistic novel views and geometrically consistent depth maps across unobserved camera viewpoints.

Reconstructing Humans and Objects in Interaction using Large Reconstruction Models

MILO reframes monocular 3D human-object interaction reconstruction as interpreting joint meshes from Large Reconstruction Models (LRMs) as non-parametric geometric scaffolds, decomposing them via multi-view segmentation, fitting a parametric SMPL-H body, and optionally aligning CAD templates to establish a new state of the art across benchmarks without contact supervision.

Recurrent Sinusoidal INRs for Efficient High-Fidelity Representation

The paper reveals that sinusoidal activations mathematically induce harmonic line-spectrum enrichment, proposing a weight-tied bias-free sinusoidal recurrence architecture with bipolar Gray-code alignment supervision that expands effective spectral support under a fixed parameter budget, achieving accelerated high-fidelity and exact discrete reconstruction across 2D and 3D tasks.

Reflection-aware generative novel view synthesis

Ref-GeNVS is the first training-free, reflection-aware generative novel view synthesis framework that treats a mirror image as two complementary viewpoints, combining mirror-gated cross-view attention with latent reflection injection to synthesize photorealistic, reflection-consistent views from single or sparse inputs.

RefracGS: Novel View Synthesis Through Refractive Water Surfaces with 3D Gaussian Ray Tracing

Addressing the non-linear optical distortion and multi-view inconsistency of underwater scene reconstruction through non-planar refractive water surfaces, this paper introduces a hybrid surface representation combining a contiguous neural height field with coarse-to-fine recursive subdivision tracing, coupled with an end-to-end differentiable 3D Gaussian ray tracing pipeline that backpropagates gradients through Snell's law to optimize the surface geometry, achieving ~15x faster training, 200+ FPS real-time rendering, and high-fidelity novel view synthesis.

RegHead: Non-Humanoid Head Blendshapes via Feed-Forward Registration

To overcome the absence of vertex correspondence and paired expression annotations for non-humanoid 3D heads, RegHead curates a dataset of approximately 20k identities under a unified semantic expression vocabulary, introducing a topology-agnostic dense stochastic anchor representation and a coarse-to-fine feed-forward registration network that predicts corresponded blendshapes without ground-truth deformation supervision and enables real-time retargeting from human tracking.

Relaxed Rigidity with Ray-based Grouping for Dynamic Gaussian Splatting

Addressing error accumulation from external motion priors and artificial over-constraint in traditional KNN-based rigid regularization, this work clusters visible Gaussians using ray blending weights and applies motion coherence and covariance eigenvalue spectral regularization, achieving consistent state-of-the-art dynamic reconstruction with zero inference overhead.

Reliability-Aware 3D Geometric Injection for Universal Person Re-identification

UniGeo decouples monocular 3D pose parameters into kinematic joint representations and introduces a consistency-aware reliability gate to dynamically inject 3D geometry as late-stage structural residuals, achieving substantial gains in clothing-change and occluded ReID while safely preventing negative transfer on clean domains.

ReViV: Reconstructing the Viewer and the View in 4D from Monocular Egocentric Video

ReViV introduces the first unified feed-forward generative framework that simultaneously reconstructs both the viewer (full-body, hand motions, gaze) and the view (scene depth, camera trajectories) in 4D from a single monocular egocentric RGB video via discrete multimodal tokenization and a Masked Generative Egocentric Transformer (MGET).

RGB-Pointmap Pretraining for Unified 3D Scene Understanding

UniScene3D proposes a unified 3D scene representation pretraining framework based on multi-view images and world-coordinate pointmaps (RGB-Pointmap), inheriting 2D foundation model priors via patch-level early fusion and enforcing cross-view geometric and grounded view alignments to achieve state-of-the-art performance across viewpoint grounding, scene retrieval, and 3D VQA.

ROAR-3D: Routing Arbitrary Views for High-Fidelity 3D Generation

ROAR-3D introduces a lightweight multi-view 3D generation framework that augments a pretrained single-view diffusion model with a token-wise view router, dual-stream cross-attention, and an orientation perturbation strategy, enabling flexible integration of arbitrary unposed multi-view inputs without camera poses.

Robust 3DGS-based SLAM via Adaptive Kernel Smoothing

Challenging the assumption that rendering sharpness governs tracking accuracy in 3DGS-SLAM, this paper introduces Corrective Blurry KNN (CB-KNN) adaptive kernel smoothing into differentiable rasterization to regularize the non-convex optimization landscape, reducing trajectory tracking errors by 19% to 35% without degrading global mapping fidelity.

Robust and Efficient Monocular 3D Gaussian SLAM for Kilometer-Scale Outdoor Scenes

Addressing the dual bottlenecks of fragile long-term tracking and prohibitive memory consumption in kilometer-scale monocular 3DGS SLAM, KiloGS-SLAM introduces a motion-adaptive hybrid tracking pipeline paired with a lifecycle-managed Gaussian mapping strategy, achieving state-of-the-art camera tracking and high-fidelity rendering across sequences exceeding 10,000 frames on a single GPU.

Rolling Shutter Camera Self-Calibration

The first target-free and IMU-free visual self-calibration framework for rolling shutter cameras, unifying continuous-time B-spline trajectory estimation and 2D pixel trajectory rectification into a dual-projection bundle adjustment to jointly estimate camera intrinsics and readout time ratio from image sequences.

SA-ResGS: Self-Augmented Residual 3D Gaussian Splatting for Next Best View Selection

Addressing the issues of biased uncertainty quantification and vanishing gradients on under-represented Gaussians during active scene reconstruction, SA-ResGS generates Self-Augmented Points (SA-Points) via single-view extrapolated triangulation for coarse physical coverage filtering, and designs an uncertainty-aware residual rasterization branch emulating skip connections to dramatically stabilize viewpoint exploration and sparse reconstruction.

SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers

Addressing the quadratic complexity bottleneck of cross-view global attention in feed-forward 3D reconstruction (F3R) transformers, SAF3R reveals head-wise heterogeneous and dynamic sparsity patterns, allocating tailored sparse kernels via offline profiling and online adaptation to achieve up to 7.3× end-to-end speedup on long sequences without retraining.

Scal3R: Learning Efficient Multi-Relative Pose Query for Scalable Online 3D Reconstruction

To tackle catastrophic drift and geometric collapse caused by global pose extrapolation in feed-forward online 3D reconstruction, Scal3R freezes the pretrained backbone and injects lightweight (~1% parameters) multi-reference relative pose query tokens via asymmetric attention, coupled with an online pose-graph optimization backend that suppresses kilometer-scale trajectory drift in only 8 hours of single-GPU tuning.

Scene Generation at Absolute Scale: Utilizing Semantic and Geometric Guidance From Text for Accurate and Interpretable 3D Indoor Scene Generation

GuidedSceneGen pioneers an absolute-scale text-to-3D indoor scene generation pipeline by leveraging a 3D metric proxy layout, multi-modal semantic/depth-guided panoramic diffusion, and collision-aware camera trajectory sampling to reconstruct fully navigable, semantically interpretable 3D Gaussian Splatting scenes without external pose estimation.

SceneHI: High-Resolution 3D-Consistent Scene Texturing with Controllable Illumination

Addressing view inconsistency, limited resolution, and missing physical illumination in complex multi-object scene texturing, SceneHI introduces a training-free framework driven by analytical pixel-to-texel mapping, high-resolution latent textures (HRLTs), patch-based refinement, and geometry-consistent shadow baking, achieving photorealistic, multi-view consistent, and relightable textures.

Seen2Scene: Completing Realistic 3D Scenes with Visibility-Guided Flow

Seen2Scene is the first flow matching approach trained directly on incomplete real-world 3D scans for scene completion and generation, introducing visibility-guided flow matching and masked sparse VAE compression that mask out unknown regions to predict high-fidelity, coherent scene geometry.

SegVGGT: Joint 3D Reconstruction and Instance Segmentation from Multi-View Images

SegVGGT introduces the first unified end-to-end framework to simultaneously perform feed-forward 3D reconstruction and 3D instance segmentation directly from unposed multi-view RGB images by coupling object queries with visual geometry transformers and introducing Frame-level Attention Distribution Alignment (FADA) with zero inference overhead.

SemCityLoc: Aerial 6DoF Localization Using Semantic 3D City Models

SemCityLoc reformulates UAV aerial 6DoF pose estimation as a "structured surface registration" problem between foundation model visual priors (semantic segmentation + monocular depth) and standardized semantic 3D city models. Through a two-stage pipeline consisting of "4D semantic cost volume coarse localization \(\rightarrow\) joint semantic-depth fine alignment via particle filtering", it improves the localization recall in urban canyon scenarios from 35.11% to 69.15% (under the 2m-2° threshold) without requiring dense radiometric reconstruction, while reducing the translation error from 9.89m to 2.62m. Furthermore, it introduces SemCityLockeD, the first real-world UAV localization benchmark with centimeter-level accuracy.

SHReg: Strictly Rotation-Equivariant Point Cloud Registration via Spherical Harmonics

Addressing the vulnerability of learning-based point cloud registration to unseen rotations caused by heuristic data augmentation or fragile local reference frames, SHReg introduces a strictly SO(3)-equivariant backbone based on spherical harmonics representation theory, simultaneously decoupling strictly rotation-invariant descriptors for matching and leveraging higher-order equivariant tensors for closed-form rigid transformation estimation from single correspondences, outperforming state-of-the-art methods across 3DMatch, 3DLoMatch, and KITTI.

SkipGS: Post-Densification Backward Skipping for Efficient 3DGS Training

SkipGS observes that the backward pass accounts for over 60% of per-iteration runtime in the post-densification phase of 3D Gaussian Splatting (3DGS) while yielding diminishing gradient returns on converged views; by introducing a view-adaptive backward gating mechanism with an auto-calibrated cumulative budget controller, it cuts post-densification training time by 42.0% and end-to-end time by 23.1% on Mip-NeRF 360 with virtually identical rendering quality.

SMG: Semantic Motion Graph for Monocular Dynamic Gaussian Splatting

SMG models monocular dynamic Gaussian deformation via a Semantic Motion Graph that restricts motion propagation within semantic groups and incorporates confidence-aware ARAP along with local rigid motion control to guide weakly constrained regions, achieving state-of-the-art novel-view synthesis and robust 3D tracking.

SOCO: Benchmarking Semantic Object Correspondence in Vision Foundation Models

Addressing the conflation of local concept recognition and object-relative spatial topology in traditional benchmarks, SOCO establishes a taxonomy-driven Semantic Object Correspondence benchmark spanning 100 categories, over 1M pairs, and structured language descriptions, exposing key bottlenecks of vision foundation models on symmetric repeated parts and LVLMs on cross-image visual matching.

Sparse auto-regressive modeling for scene generation from multi-view images

Addressing the challenge of full 3D scene completion from sparse, unconstrained views without 3D ground truth, SPAR3S proposes a sparse voxel-aligned 3D latent masked autoregressive framework combining token-wise diffusion with differentiable 3D Gaussian Splatting to achieve high-fidelity novel view synthesis and unobserved geometry extrapolation.

Sparse-View Surface Reconstruction using Gaussian Splatting through High-Confidence Depth Propagation with Normal Priors

Addressing ambiguous geometry and surface tearing caused by Gaussian discreteness under sparse views, this paper identifies high-confidence depth anchors via multi-view geometric and photometric consistency, iteratively propagates depths outward guided by normal priors, and introduces an abnormal depth edge-aware smoothing regularizer to reconstruct continuous, high-fidelity surfaces.

SPECSIA: Stylization Dataset for Novel-View Enhancement in Drawing-based 3D Animation

This paper proposes the SPECSIA-15K paired stylization dataset (14,980 pairs of artifact-laden projections and clean targets) and DraViE, a lightweight plug-and-play post-correction module. By replacing sample-level frame-by-frame optimization with a data-level pretraining prior, it effectively eliminates projection artifacts in novel views for drawing-driven 3D animation while maintaining the original drawing style.

SpectralSplats: Robust Differentiable Tracking via Spectral Moment Supervision

This work shifts the supervision target of 3DGS differentiable tracking from the spatial domain (pixel-wise photometric error) to the frequency domain (spectral moments). Combined with a frequency annealing schedule strictly derived from the principle of phase wrapping, the method enables Gaussian assets to obtain non-zero directional gradients even when completely non-overlapping with the target at initialization, smoothly "flowing" to the correct pose—thereby enabling robust tracking without manual alignment or category priors.

Stable and Scalable Bundle Adjustment of Holistic 3D Structures

This paper introduces a unified bundle adjustment framework that incorporates higher-order geometric relations (planes, vanishing points, wireframes) by treating groups as camera-like non-eliminated variables and formulating constraints as 2D reprojection residuals, preserving the block-diagonal Schur complement efficiency of classical point BA while yielding substantially richer and more accurate 3D geometry.

StereoGS: Sparse-View 3D Gaussian Splatting via Stereo Priors

StereoGS addresses the inherent scale ambiguity and cross-view inconsistency in monocular depth priors of sparse-view 3DGS by constructing virtual stereo camera pairs, introducing stereo depth regularization, and applying gradient-aware opacity decay, achieving state-of-the-art novel view synthesis on four datasets: LLFF, DTU, Mip-NeRF360, and Blender.

Stitched Embeddings: A Unified Latent Space for 3D Garments and 2D Patterns

StEm-Net establishes the first simulation-free, fully differentiable bidirectional latent space that unifies 3D draped clothing and 2D sewing patterns via BoxMesh proxies, delivering state-of-the-art pattern recovery alongside gradient-driven test-time refinement and instant 3D interactive editing.

StructureGS: Structure-aware Gaussian Splatting for Articulated Object Reconstruction

StructureGS tackles the entanglement between geometry and motion caused by purely photometric supervision in articulated object reconstruction by introducing explicit oriented bounding boxes (OBBs) and canonical 3D Gaussian primitives, regularized via differentiable part fitting and SAT-based part contact losses to enable clean part boundaries, precise kinematics, and robust sparse-view reconstruction.

StyleFusion360: View-Consistent Head Stylization via Adaptive Style Modulation

Addressing the challenge of balancing fine geometric preservation with zero-shot generalization across unseen styles, StyleFusion360 builds upon a 3D-aware diffusion backbone with decoupled appearance branches and introduces a two-stage Style Fusion Attention with adaptive key modulation, delivering 360-degree view-consistent and identity-preserving head stylization without per-style retraining.

SubSplat: High-Resolution Pixel-aligned 3DGS via Sub-pixel Gaussian Reparameterization

SubSplat introduces a Sub-pixel Gaussian Reparameterizer (SPGR) coupled with deformable multi-view feature aggregation, which subdivides each grid-anchored primary Gaussian into fine-grained 3D primitives with conserved opacity redistribution directly from low-resolution inputs, eliminating quadratic backbone computation while achieving real-time, high-fidelity rendering.

SynCity 3000: Bootstrapping Scene-Scale 3D Diffusion

SynCity 3000 introduces a two-stage framework for large-scale 3D scene generation that first produces an ultra-high-resolution dimetric scene template via overlapping latent window denoising, and then converts it into seamless, arbitrary-scale 3D Gaussian Splats through a convolutional 3D diffusion model fine-tuned on procedurally bootstrapped synthetic scenes.

TabletopGen: Tabletop Scene Generation and Interactive Simulation for Robotic Manipulation

TabletopGen is a training-free, automated 3D tabletop scene generation and interactive simulation engine that decouples layout recovery into generative instance extraction and two-phase pose/scale alignment (DRO rotation optimization and TSA top-view spatial alignment), synthesizing physically plausible, collision-free tabletop environments that enable zero-shot real-to-sim-to-real robotic policy transfer.

Temporally Aware Densification for Dynamic 3D Gaussian Splatting

Addressing the failure of dynamic 3DGS to densify short-lived and highly dynamic Gaussians under static densification schemes, this paper introduces visibility-weighted gradient accumulation (VAD), lifespan-aware adaptive thresholding (TAT), and focused temporal offset warping (TOW), delivering substantial gains in dynamic reconstruction quality and high rendering frame rates.

TetraSDF: Analytic Isosurface Extraction with Multi-resolution Tetrahedral Grid

TetraSDF introduces a continuous piecewise affine (CPWA) neural SDF framework combining a multi-resolution tetrahedral grid with a shallow ReLU MLP, leveraging the inherently affine nature of barycentric interpolation to achieve exact analytic zero-level set isosurface extraction while preserving high-frequency shape fitting capability.

The Dynamic Prior: Understanding 3D Structures for Casual Dynamic Videos

This paper introduces the training-free Dynamic Prior (\(D_{\text{prior}}\)), which leverages Vision-Language Models (VLMs) via chain-of-thought reasoning to identify moving objects and generates temporally coherent masks using Sa2VA with SAM2, serving as a plug-and-play filter that significantly enhances camera pose optimization, consistent depth reconstruction, and 4D trajectory estimation in casual in-the-wild videos.

TIDES: Time-Derivative Event Simulation via Deformable Reconstruction

Addressing the pervasive timestamp batching and occlusion distortion caused by discrete frame differencing in conventional event simulators, TIDES leverages dynamic 4D Gaussian splatting with forward-mode time-derivative rasterization to analytically solve continuous-time threshold crossings directly from explicit 3D dynamics, combined with occlusion-risk adaptive time-stepping and a physical tile readout arbiter.

TimeWalker: Personalized Neural Space for Lifelong Head Avatars

Addressing the core limitation of conventional head avatar pipelines that only model momentary static captures or short video clips, TimeWalker introduces a personalized neural linear combination space and Dynamic 2D Gaussian Splatting (DNA-2DGS) to reconstruct and explicitly animate lifelong 3D head avatars with robust identity consistency across ages from unstructured in-the-wild photo collections.

TopoGS: Planar Reconstruction via Topology-Aware 3D Gaussian Splatting

TopoGS couples 2D region adjacency graphs with 3D Gaussian Splatting via a tri-consistency optimization framework incorporating uncertainty-aware topological alignment, filtering view-dependent spurious adjacencies and producing sharp, topologically connected 3D planar models.

Towards Alias-Free 4D Gaussian Representations with Motion-Aware Filtering

To tackle severe aliasing and blurring in dynamic 4D Gaussian representations during novel-view zoom-in and zoom-out operations, this paper proposes a motion-aware 3D smoothing filter that models the joint density of time and focal-to-depth ratio via non-parametric kernel density estimation, enabling scale-consistent dynamic rendering.

Towards in-the-wild Egocentric 3D Hand-Object Pose Estimation

This work constructs the first in-the-wild egocentric hand-object contact dataset, EPIC-Contact, using a low-cost annotation pipeline of "hand contact area \(\to\) bijective transfer to object \(\to\) optimization fitting." It also proposes HOPformer, an end-to-end network that injects pre-trained hand priors into object features via cross-attention to jointly regress the 3D poses of both hands and the object in a single forward pass. This improves the success rate from 76.2% to 82.4% on ARCTIC, while nearly doubling the success rate and reducing the contact deviation by 75% on in-the-wild data.

Triangle Splatting SLAM

The first dense RGB-D SLAM system that adopts a differentiable triangle soup as its sole underlying map representation, utilizing analytic camera pose Jacobians, equilateral geometric regularization, and online restricted Delaunay triangulation to deliver superior 3D geometric reconstruction and real-time mesh editing compared to 3D/2D Gaussian Splatting baselines.

UBone3D: Physics-Rectified Conditional Flow Matching for Anatomical 3D Shape Completion from Ultrasound

UBone3D decouples anatomical plausibility from imaging physics consistency by combining a high-fidelity CT-trained conditional flow matching generative prior (BoneFM) with a differentiable acoustic physics proxy (USimNet), dynamically steering sampling trajectories at test time via explicit analytical gradients to achieve accurate anatomical 3D shape completion from severely artifact-laden and partial ultrasound point clouds.

UniFusion: Sparse-View 4D Reconstruction via Unified Spatio-temporal Depth Alignment

UniFusion introduces a unified spatio-temporal depth alignment framework that models depth deformation via low-rank residual neural fields, jointly resolving cross-view and cross-temporal inconsistencies without foreground segmentation or point tracking to provide robust geometric priors for 4D Gaussian splatting.

UniPart: Towards Zero-shot Language-Grounded 3D Part Segmentation for Embodied Interaction

UniPart proposes a lightweight feed-forward cross-modal 3D Transformer with layer-wise additive text conditioning alongside LangPart-1M (160K+ objects and 8M text-part pairs generated via multi-view consistent merging) and a human-verified LangPart-4K subset, achieving real-time (0.4s) zero-shot 3D part segmentation and real-world robot grasping without multi-view backprojection at inference.

UniSim-SLAM: Feed-Forward SLAM with Unified Sim(3) Optimization

Addressing the fundamental trade-off between two-view low-latency tracking and multi-view constraint richness in feed-forward SLAM, UniSim-SLAM formulates a unified Sim(3) multi-level factor graph that jointly optimizes global keyframe and submap poses, reducing trajectory error by 38.5% on TUM RGB-D and 45.9% on 7-Scenes under uncalibrated settings.

UniTriSplat: A Unified 3D Gaussian Splatting Framework with Uniform Spherical Rasterization for Universal Cameras

This work shifts the rasterization of 3D Gaussian Splatting from "planar projection for a specific camera class" to the "HEALPix equal-area grid on the unit sphere". By utilizing a single camera-agnostic spherical rasterizer to universally support perspective, fisheye, and panoramic cameras, it simultaneously eliminates polar oversampling in equirectangular projections and achieves stable cross-camera generalization.

Unsupervised Pixel-Level Semantic Left-Right Understanding of In-the-Wild Images

Addressing the fundamental challenge of pixel-level semantic left-right prediction on in-the-wild images where 3D geometry is absent, Pix2LR introduces an unsupervised cross-modal framework that swaps the order of prediction and surface aggregation, effectively transferring geometric chirality priors from limited 3D shapes to dense pixel predictions on open-vocabulary natural images.

URHead: A Unified UV-Space Representation for Joint Mesh–3DGS Optimization in Head Avatars

URHead introduces a unified UV-space parameterization to jointly optimize parametric facial meshes and 3D Gaussian Splatting (3DGS) via shared attribute maps and error-driven adaptive sampling, achieving photorealistic facial details while maintaining stable topological animation controllability.

Vitality-Aware Compression for Efficient Image-to-Shape Diffusion Transformers

For image-to-3D Diffusion Transformers (DiTs), this paper introduces a "layer vitality" metric based on Earth Mover's Distance to measure each layer's contribution to geometric synthesis. Based on this, a pipeline of structured pruning, layered mixed-precision quantization, and targeted distillation that fine-tunes only the lowest-vitality layers is developed. This compresses backbones by up to 66% on state-of-the-art models like Step1X-3D and Hunyuan3D while maintaining geometric fidelity.

VOCA: Visual Odometry with Codec Awareness

VOCA leverages the built-in motion vectors from the H.264 video encoding process as an initialization prior for KLT optical flow tracking. Built upon the Basalt framework, it achieves high-accuracy causal visual odometry on videos compressed by up to 100x, comprehensively outperforming existing methods on the EuRoC, TUM-VI, and MSD datasets.

Walking in the Implicit: Interactive World Exploration via Neural Scene Representation

NeuWorld replaces the "rollout state" of interactive world exploration from an ever-growing sequence of video frame latent variables with a set of fixed-length, renderable, and locally anchored neural implicit scene (NIS) tokens. This allows the diffusion model to transition only within this compact state space, while a frozen decoder renders observations based on queried camera poses. Consequently, without leveraging any pre-trained video foundation models and training entirely from scratch, NeuWorld achieves excellent long-term pose consistency, loop-revisit consistency, and inference efficiency.

Weather-Conditioned Depth Anything

To tackle catastrophic failures and feature entanglement in monocular depth foundation models under adverse weather conditions (rain, snow, fog, night), this paper presents DA-W (Weather-Conditioned Depth Anything), which disentangles weather style via a cross-domain contrastive Style Filter and modulates a frozen Depth Anything backbone through parameter-efficient, zero-initialized AdaLN-Zero layers without hurting clean-scene generalization.

What VGGT Knows About Overlap: Probing Geometric Foundation Models for Co-Visibility

By freezing the geometric foundation model VGGT and attaching a lightweight layer-wise mixture-of-experts (MoE) head (~7.5M parameters), Co-VGGT predicts calibrated co-visibility probabilities from RGB alone, outperforming the human annotation baseline, boosting pairwise IoU on Co-VisiON by over 25% compared to prior SOTA, and revealing layer 17 as an emergent negative anchor specialized in rejecting non-overlapping pairs.

When 3D Gaussian Splatting Recovers Real Surfaces

This paper develops a mathematical framework based on a first-hit rendering abstraction to investigate the geometric identifiability of 3D Gaussian Splatting (3DGS), proving that geometric misalignment forcefully converts spatial textures into high-frequency angular signals via parallax, and establishing an intermediate spherical harmonic capacity window where true surface recovery is strictly preferred over opaque billboard collapse.

White Aggregation and Restoration for Few-shot 3D Point Cloud Semantic Segmentation

To tackle performance instability caused by Farthest Point Sampling (FPS) and attention collapse induced by distributional mismatch between prototypical tokens and support point cloud features, WARM wraps cross-attention with ZCA whitening and coloring restoration, achieving new state-of-the-art results on S3DIS without complex decoders.

WildDepth: A Multimodal Dataset for 3D Wildlife Perception and Depth Estimation

WildDepth establishes the first synchronized RGB-LiDAR field benchmark covering 29 animal species and 202k curated frames across diverse natural habitats, revealing severe scale degradation in state-of-the-art monocular depth estimation models at long range and demonstrating that sparse LiDAR anchoring coupled with deformable Gaussian Splatting enables accurate metric 3D/4D reconstruction in the wild.

WildSplat: Feedforward Gaussian Splatting from Unposed In-the-Wild Images

WildSplat is the first feedforward 3DGS framework for unposed in-the-wild images: a geometry branch (DINOv2 encoder + 24 alternating-attention layers) extracts appearance-invariant content features and jointly predicts camera poses, while an appearance branch uses [CLS] pre-modulation plus cross-attention to inject reference-image appearance before predicting conditioned spherical-harmonics colors. Joint multi-reference supervision prevents geometry-appearance entanglement; on MegaScenes/Phototourism at 4/8/12 views it comprehensively beats AnySplat, WorldMirror, and optimization-based methods (Phototourism 12-view PSNR 21.47), and supports single-pass appearance editing and interpolation.

WilLaGS: Latent-Conditional 3D Appearance Fields for Robust Gaussian Splatting In-the-Wild

WilLaGS learns a continuous global appearance space with a β-VAE, generates a Tri-Plane 3D appearance field from its latent code to modulate local illumination, and masks transient occlusions using perceptual differences between an EMA teacher and real images, reaching 25.84 dB on Photo Tourism's Sacre Coeur while providing 58 FPS rendering, appearance interpolation, and random sampling within the same scene.

2D Features Are All You Need for 3D Shape Understanding

MeshFM fits corrected 2D foundation-model features into continuous 3D teacher fields, then trains a rotation-augmented feedforward predictor to reproduce them without manual 3D annotations, supporting segmentation, correspondence, and deformation with the same features and achieving 0.549/0.539 semantic mIoU on original/rotated PartObjaverse-Tiny shapes.

3D Gaussian Splatting Compression with Object Scalability

RecastGS reorganizes pretrained 3DGS into a hierarchy with object-specific quality choices, and LayeredCGS compresses it for progressive decoding; on kitchen near 2.3 MB, ROI PSNR rises from 19.65 to 21.81 dB, at the cost of lower background and full-image quality.

3D Gaussian Texture for Real-time Mesoscale Appearance Synthesis and Rendering

The paper organizes 3D Gaussians into regular texture layers at similar heights above a surface, enabling real-time exemplar-based synthesis, transfer to new meshes, and basic relighting of structures such as grass, durian spikes, and wicker gaps; it improves SIFID over NeRF-Texture in all four comparisons, with reported times of 3.28–4.98 ms in Figure 4.

3D-ReGen: A Unified 3D Geometry Regeneration Framework

MeshReGen represents coarse meshes, incomplete point clouds, and masked geometry as VecSet conditions, concatenates their tokens with noisy target tokens, and learns image-guided 3D regeneration from automatically constructed degradation pairs, reaching CD 0.0081 and F-score 0.4913 on four-view GSO reconstruction while also supporting enhancement and local editing.

3DGS3: Joint Super Sampling and Frame Interpolation for Real-Time Large-Scale 3DGS Rendering

Instead of rasterizing every output frame at high resolution, 3DGS3 jointly recovers spatial and temporal resolution through analytical-gradient-guided super sampling and lightweight frame interpolation, increasing 3DGS throughput from 17.35 to 96.47 FPS at 4096×2160 on Mip-NeRF 360 while achieving 49.22 dB PSNR against high-resolution reference renderings.

4D-VGGT: A SpatioTemporal Foundation Model for Dynamic Scene Geometry Estimation

4D-VGGT organizes inputs with an adaptive view-time grid, learns spatial and motion cues through parallel cross-view global and cross-time local fusion, and routes them to five geometry heads, achieving a camera-pose ATE of 0.017 on Bonn while using ablations to demonstrate distinct spatial and temporal contributions.

Abstract the Layout, Focus the Detail: A Dual-Granularity Representation Framework for Zero-Shot 3D Visual Grounding

The framework recasts a 3D scene as an abstract layout for spatial reasoning and object-centric patches for appearance verification, enabling an off-the-shelf VLM to select targets in a spatial-first, visual-second workflow and achieve 45.0% [email protected] on ScanRefer and 63.2% accuracy on Nr3D without GT object classes.

ActiveStructure: Plane Scene Graph-Guided Active 3D Gaussian Splatting

ActiveStructure augments active 3DGS mapping with a persistent plane scene graph, uses relational anomalies and localized 3D gaps to select views, and turns established walls from exploration attractors into suppressors, achieving 97.47% coverage and 32.13 dB PSNR on Replica within a fixed 2000-frame budget.

Ada-VNNs: Adaptive Equivariance for Vector Neural Networks

Ada-VNNs attach a residual branch with relaxed coordinate-wise weight sharing to a strictly equivariant vector-neuron backbone, raising VNN's 10° registration recall from 8.87% to 70.31% on approximately continuously self-symmetric ShapeNet categories, while introducing task-dependent trade-offs rather than universal improvements.

AdaptiveSplat: Texture Aware Controllable 3D Gaussian Allocation for Feed-Forward Reconstruction

AdaptiveSplat uses local texture complexity to decide where to prune 3D Gaussians, then predicts updated attributes through a mask-conditioned head, enabling budget-controlled reconstruction without test-time scene optimization and reaching 20.740 dB PSNR on RE10K after pruning 80% of the primitives.

Aggregating Cross-Domain Knowledge via Learnable Tokens for Multi-Teacher Distillation

ACTok places channel-concatenated learnable tokens and alternating cross-space/intra-space interactions after a student backbone to consolidate CLIP, MAE, and DINO knowledge using ImageNet-1K, reaching 94.4% on Caltech101 with ViT-B without universally outperforming the strongest teacher or distillation baseline.

AnyGround3D: Towards Grounding Any 3D Object in the Wild via 2D-to-3D Lifting

AnyGround3D automatically lifts monocular images from sources such as COCO and KITTI into 3D grounding supervision containing scene point clouds, complete-object 3D boxes, and natural-language descriptions, then distills DINOv2 semantics into its 3D encoder; it obtains 66.76/31.34 [email protected]/0.5 on the 3EED union split and 62.71/54.79 on the Mono3DRefer test set without target-specific fine-tuning.

AutoPhyX: Automatic Text-Condition Physics Property Generation

AutoPhyX introduces the first text-conditioned feed-forward framework for 3D physical property prediction, leveraging volume-rendered OpenCLIP voxels and FiLM cross-modal modulation to regress dense Young's modulus, Poisson's ratio, and density fields in a single forward pass, resolving the fundamental physical ambiguity of purely visual methods.

AVSplat:Dense-View Feed-Forward 3D Gaussian Splatting with Assist-View Preconditioning

Addressing dense-view reconstruction degradation in pose-free feed-forward 3D Gaussian Splatting caused by attention dispersion and naive voxel averaging, AVSplat introduces Assist-View Preconditioning and Occupancy-guided Voxel Fusion with adaptive temperature, restoring positive view scaling as input views increase.

BioMTBee: Biologically Constrained Multi-View Template-Based 3D Reconstruction of Bumblebee

BioMTBee fits a micro-CT-derived bumblebee template to multi-view video using filtered 3D poses, silhouettes, and biological constraints, reducing mesh-keypoint error from MAMMAL's 0.189 mm to 0.152 mm while increasing silhouette IoU from 0.696 to 0.758 on the DBS test split.

BLASt3R: Bundle Adjustment of Any Image Set with Multi-View Matching and Monocular Depth Priors

BLASt3R introduces a hybrid 3D reconstruction system that combines streaming coarse-to-fine multi-view feature matching with adjustable monocular depth priors, jointly solving camera poses, tracks, and depth deformation parameters within a unified GPU-accelerated regularized bundle adjustment pipeline across both offline SfM and online VSLAM setups.

Boosting 6D Object Pose Estimation via Monocular Depth Cues

MDC-Net couples monocular-depth calibration, geometric inlier selection, and recurrent pose refinement, achieving 66.9% mean AR across seven BOP datasets with one pose hypothesis and no real depth at inference, although several auxiliary results and training descriptions are inconsistent.

Boosting Correspondence Learning with Structure-Aware Estimator

A structure-aware estimator replaces independent weighted least squares at the end of correspondence-learning networks, using learned inlier relationships to improve DeMatch's YFCC100M AUC@5° from 30.95 to 36.58, with important limitations on indoor–outdoor transfer and correspondence-pruning architectures.

Bootstrapping Articulated 3D Reconstruction from 2D Image Collections

BAT3R initializes a 3D predictor from a single canonical-pose mesh, then fits articulations from 2D images, re-renders them, and trains iteratively, reducing horse RMS Chamfer Distance on Animodel-Points from 7.88 cm to 5.65 cm, while remaining above the 4.30 cm achieved by fully 3D-supervised DualPM.

BRepFacetGen: Reverse Engineering B-Reps By Generative Face Segmentation

To overcome topological instability and structural ambiguity in traditional deterministic point-cloud CAD reverse engineering, BRepFacetGen formulates CAD face decomposition as a generative latent variable problem conditioned on continuous geometry, coupling pretrained 3D foundation models with flow-matching label synthesis to deliver superior geometric and topological fidelity in B-Rep reconstruction.

Bridge-UniPS: Bridging Calibrated Photometric Stereo toward Universal Photometric Stereo

Bridge-UniPS introduces a bridging paradigm that connects universal photometric stereo (UniPS) to calibrated photometric stereo (CPS) via a task-driven lighting adapter that translates unconstrained multi-illumination observations into CPS-compatible bridge image–illumination pairs without intermediate supervision, achieving state-of-the-art normal estimation accuracy by leveraging the physical priors frozen in pretrained CPS networks.

CascadeProto: Cascaded Cross-Modal Prototype Purification via Entropy-Aware Learning for Few-Shot 3D Point Cloud Segmentation

To tackle background noise contamination and semantic scarcity in few-shot point cloud prototypes, CascadeProto proposes cascaded entropy-aware purification alongside decoupled cross-modal alignment, achieving new state-of-the-art accuracy on S3DIS and ScanNet without any pre-training.

CAST3D: Customizing Arbitrary 2D Assets into 3D World

CAST3D presents a training-free framework for customized 3D composition, bridging 2D assets and 3D generation through 3D layout hinting, stochastic trajectory manipulation, and connectivity-based pruning to synthesize coherent, faithful 3D objects under textual guidance.

Category-Level Articulated Object Pose Estimation via Pose–Shape Hypothesis Generation and Verification

To tackle severe occlusion-induced pose ambiguity and sensor depth noise in single-view articulated object pose estimation, this paper introduces a generation-and-verification framework that jointly samples coupled pose-shape hypothesis pairs via normalizing flows, recovers clean part-level geometries, and prunes geometrically invalid hypotheses using base-anchored joint alignment and continuous Signed Distance Function (SDF) consistency, significantly outperforming existing single-view baselines and sequence trackers.

Closing the Capacity–Convergence Gap: Globally Optimal Configuration of Implicit Neural Representations

OptiINR reformulates layer-wise activation selection, initialization scaling, and learning rate configuration as a mixed-variable black-box global optimization problem, using product-kernel Gaussian processes and Matheron's rule-accelerated Empirical Expected Improvement to bridge the capacity–convergence gap in coordinate-MLPs.

Consistent Monocular Depth Estimation with Contact Region Boundary-Aware Refinement

To eliminate pervasive and erroneous depth discontinuities at object contact regions in monocular depth estimation, this paper introduces a contact boundary-aware estimation and refinement framework that distills physical contact boundaries using vision-language semantics, fuses them into the depth decoder via multi-scale structural priors, and enforces multi-order geometric constraints, significantly resolving contact surface gaps with only 4,000 synthetic images and supporting controllable interactive refinement.

Controllable Generative Reference for Stereo Image Compression via Reliability-Aware Gating

Addressing the vulnerabilities of traditional stereo image compression in out-of-view (OOV) regions and the propagation of heavy quantization artifacts, this work leverages a pre-trained latent diffusion model steered by joint semantic-spatial control to synthesize a high-quality target-view reference, adaptively regulated via reliability-aware gating within the entropy model to achieve state-of-the-art rate-distortion performance.

Cube-Splat: High-Fidelity 360° Gaussian Splatting SLAM via Cubemap Factorization and Adjoint-Consistent Optimization

Addressing the narrow field-of-view, severe trajectory drift, and non-frontal rendering voids of conventional pinhole 3DGS-SLAM, Cube-Splat factorizes 360° panoramic imagery into four virtual pinhole cubemap faces sharing a single optical center, aggregating multi-face photometric gradients into a single primary pose via Lie-algebra adjoint mappings for robust tracking and omnidirectional high-fidelity reconstruction.

DA-F2F: Domain-Adaptive Object Detection with Feature-to-Feature Modulation and Alignment

Addressing the computational burden, pixel artifacts, and pseudo-label instability of traditional image-to-image translation in domain adaptive object detection, DA-F2F dynamically modulates source features with target style statistics (SFM) in latent space and applies continuous soft-weighted proposal adversarial alignment (SPA), achieving robust cross-domain alignment and new state-of-the-art detection accuracy.

DASAM3D: A Unified Foundation Model for Enhanced 3D Scene Reconstruction and Segmentation

DASAM3D utilizes SAM 3D's high-quality per-object Gaussian primitives exclusively as a 3D geometric layout provider to guide the geometry refinement of Depth Anything 3 (DA3) across three differentiable stages without fine-tuning either foundation model, preserving DA3's photorealistic appearance while resolving floating Gaussians and distorted surfaces.

Detect by Track: Making Detector-Free Matcher Trackable

Addressing the core multi-view inconsistency and track fragmentation inherent to implicit keypoint selection in detector-free matchers, this paper proposes Detect-by-Tracking (DeT), an inference-time steering mechanism that enables off-the-shelf detector-free matchers to directly generate sub-pixel connected tracks without any retraining.

DeWorldSG: Depth-Aware 3D Semantic Scene Graph Generation via World-Model Priors

DeWorldSG proposes a framework for generating spatiotemporally robust 3D semantic scene graphs (SSGs) from RGB-D sequences: SAM-mask-guided dual-domain depth filtering estimates an instance-level probabilistic 3D Gaussian distribution for each object, while the V-JEPA 2 world model accumulates cross-frame relational evidence and refines predicate edges through uncertainty-aware fusion, improving relationship recall by 77.4% and predicate recall by 23.2% on 3DSSG.

DINOv3D: 2D-3D Joint Optimization for Unified Spatial Understanding

DINOv3D presents a homologous teacher-student joint optimization framework based on DINOv3, where a frozen teacher anchors native 2D visual priors into a 3D Gaussian regularization field while a LoRA-adapted student ingests long-context 3D geometric consistency via shifted-window dual-memory attention and ray-depth-semantic alignment, establishing state-of-the-art results across 3D scene understanding benchmarks with only 10% trainable parameters and reciprocally enhancing 2D spatial features.

DiTex4D: Direct Text-Driven 4D Generation with Structured Latent Diffusion

DiTex4D builds a native text-driven 4D generation framework upon the 3D foundation model TRELLIS, eliminating intermediate video generation via mixed-4D RoPE and spatial-keyed correlated noise injection to natively support both 4D generation from scratch and mesh animation.

DLGStream: Dynamic Language-embedded Gaussian Splatting for Open-vocabulary Enabled Free-viewpoint Video Streaming

DLGStream introduces a dual-opacity language Gaussian representation and a deformation field based on temporal feature interpolation, embedding streamable spatiotemporal language features while reconstructing high-quality free-viewpoint video from multiview dynamic scenes, using an average frame size of 43 KB to enable real-time 4D open-vocabulary queries and frame interpolation.

Don’t Mask Out the Background! Natural-Light Photometric Stereo via Illumination Reconstruction

Addressing the severe ill-posedness of uncalibrated natural-light photometric stereo (NaPS) with unknown lighting, this paper discards the standard practice of masking out image backgrounds, leveraging background cues captured under flexible rigid camera-object motions to explicitly reconstruct a spatially aware, near-field 3D Gaussian Splatting (3DGS) illumination emitter, and subsequently optimizes surface normals, Disney BRDF reflectance, and depth via physics-based inverse rendering.

Don’t Starve the Boundaries: Boundary-Constrained Label Propagation for Weakly Supervised 3D Segmentation

Addressing the "boundary starvation" issue where limited supervision fails to provide reliable boundary guidance in weakly supervised 3D point cloud segmentation, this paper introduces Bound3D—a framework that projects multimodal 2D geometric discontinuities to identify 3D boundaries, generates high-purity pseudo-labels via manifold-constrained propagation and farthest angle sampling, and optimizes interior and boundary regions through position-aware divide-and-conquer supervision.

DPGS: A Diffusion-Prior Guided Framework for Large-Scale 3D Gaussian Splatting Reconstruction

Addressing facade blind spots, weak-texture geometric collapse, and parallel rendering load imbalances under restricted nadir-view aerial flight missions, DPGS introduces foundation model dense initialization, track-length-weighted graph partitioning, and a reference-conditioned diffusion completion pipeline to deliver geometrically complete, high-fidelity large-scale 3DGS scenes.

DualResPS: Dual-Resolution Photometric Stereo Using a Frame-Event Hybrid Camera

DualResPS presents a dual-resolution photometric stereo pipeline utilizing a coaxial frame-event hybrid camera under a continuously rotating light source, integrating sparse long-exposure frames with high-temporal-resolution events via coordinate-based neural inverse rendering to recover high-resolution surface normals at only 18.1% of the data rate.

EGGS: Explicitly Granular 3D Gaussian Splatting via Luma-Aware and Volume-Preserving Attribute Factorization

Targeting the massive storage footprint caused by the explicit attribute parameters in 3DGS, EGGS introduces luma-based spherical harmonics DC compression and volume-preserving scale factorization to reduce per-primitive explicit attributes to as few as 9 dimensions, compressing gigabyte-scale scenes into a few megabytes while sustaining over 140 FPS real-time rendering.

EgoSim: Egocentric World Simulator for Embodiment Interaction Generation

EgoSim introduces a closed-loop egocentric world simulator that couples a geometry-action-aware video diffusion model with an updatable 3D point cloud world state, generating spatially consistent, continuous interaction videos with seamless cross-embodiment transfer.

EPO: Boosting 3D Foundation Models with Edge-based Pose Optimization

This paper proposes EPO, a geometric optimization framework that does not require explicit feature tracks. It utilizes Canny edge maps and distance transform fields to construct a differentiable edge reprojection loss, refining the camera parameters and depth maps output by 3D foundation models through a two-stage adaptive optimization. It achieves or exceeds the geometric accuracy of traditional bundle adjustment (BA) methods within seconds on consumer-grade GPUs.

ETCH-X: Robustify Expressive Body Fitting to Clothed Humans with Composable Synthetic Data

ETCH-X establishes an "undress first, then dense fit" modular paradigm that unifies local SE(3)-equivariant tightness vector prediction with implicit dense neural fields, enabling scalable training on composable synthetic datasets (CLOTH3D garments, AMASS poses, and InterHand2.6M gestures) for expressive and robust SMPL-X body fitting under diverse clothing, complex motions, and partial point cloud scans.

EventSpecPS: Photometric Stereo with Multispectral Reflectance Using an Event Camera

EventSpecPS introduces the first single-stage event-based multispectral photometric stereo system, leveraging spectrally-angularly multiplexed illumination and a constant-light bias to decouple geometry from spectral reflectance and resolve scale ambiguity, achieving low-bandwidth, robust joint recovery of surface normals and absolute multispectral reflectance.

Fast Spatial Memory with Scalable Elastic Test-Time Training

To resolve catastrophic forgetting and fast-weight drift caused by unconstrained plasticity in Large Chunk Test-Time Training (LaCT) on dynamic sequences, this paper introduces Large Chunk Elastic Test-Time Training (LaCET) with Fisher-weighted elastic consolidation around a streaming-EMA anchor, establishing an efficient and scalable 4D spatial memory model (FSM).

FFAvatar: Feed-Forward 4D Head Avatar Reconstruction from Arbitrary Images

FFAvatar presents a feed-forward 4D head avatar reconstruction framework that incrementally reconstructs high-fidelity, controllable 3D Gaussian avatars from one or multiple arbitrary portrait images using alternating attention for identity disentanglement, a sparse-to-dense canonical alignment strategy, and a motion-aware dynamic refinement module.

Flash-Refine: Frustum-Guided Local Incremental Learning for Efficient 3D Gaussian Splatting Completion

Presents Flash-Refine, a practical in-situ completion framework for static 3D Gaussian Splatting (3DGS) assets that identifies defective regions via multi-view frustum consensus and adaptive depth priors while locking background gradients, completing degraded geometry in 2–4 minutes without historical training images.

F⁴Splat: Feed-Forward Predictive Densification for Feed-Forward 3D Gaussian Splatting

Addressing the primitive redundancy and lack of budget control in conventional feed-forward 3D Gaussian Splatting caused by rigid pixel- or voxel-based allocation, F⁴Splat introduces a feed-forward predictive densification framework that dynamically allocates Gaussians based on predicted densification scores, enabling explicit Gaussian budget control without retraining and cutting primitive counts by over 70% while improving rendering fidelity.

G2FM: A Geodesic Flow Matching Framework with Geometric Prior for Category-Level 9-DoF Pose Estimation

Addressing the slow multi-step sampling and rotation orthogonality corruption of diffusion models in category-level 9-DoF pose estimation, G2FM introduces continuous geodesic flow matching on the \(SO(3) \times \mathbb{R}^3 \times \mathbb{R}^3\) product manifold alongside training-free tangent-space Chamfer guidance, attaining 31.5 FPS real-time estimation with only 3 ODE solver steps.

Geometric Context Transformer for Streaming 3D Reconstruction

GCT draws on classical SLAM principles to decompose streaming spatial context into anchor context, a local pose-reference window, and lightweight trajectory memory via Geometric Context Attention (GCA), slashing per-frame token growth by ~80× and achieving state-of-the-art streaming reconstruction at 20 FPS across long sequences exceeding 10,000 frames.

Geometric Distillation from Rectified Stereo: Leveraging Epipolar Cues for Monocular Depth Estimation

To prevent multi-view foundation models from suffering architectural attention collapse and losing scale priors during single-view inference, EpiDistill transfers multi-view epipolar geometry and metric scale priors into a monocular Transformer using depth-guided epipolar attention and rectified stereo tokens, significantly advancing zero-shot metric depth estimation.

Geometric Probing for Isotropic Optimization Manifold in Sparse-View 3D Gaussian Splatting

Addressing the severe directional anisotropy of the 3DGS loss surface under sparse views which triggers floaters and splat collapse, StableGS steers optimization toward an isotropic manifold via Hessian-informed Attribute-Space Probing and Viewpoint-Space Probing without external priors, reducing anisotropy by 74% and advancing SOTA across LLFF, DTU, and Mip-NeRF360.

Geometry-Propagated Gaussian Splatting for Aerial Sparse Novel View Synthesis

Addressing the dual bottlenecks of SfM feature matching collapse and extensive unobserved areas in aerial sparse-view settings, GeoProp-GS proposes Depth-guided Geometric Initialization (DGI) and Anchor-constrained Gaussian Optimization (AGO), which densify and complete geometry before optimization and propagate reliable geometric priors into under-supervised regions via an anchor-residual parameterization to achieve state-of-the-art sparse-view rendering.

GlassGS: Geometry and Concept-Aware 3D Gaussian Splatting for Reflective Enclosures

To tackle severe multi-view inconsistencies caused by high specular reflections on glass display cases, GlassGS presents an iterative two-stage framework coupling geometry-consistency perception with reflection-free concept learning, achieving artifact-free, high-fidelity 3D reconstruction on a dedicated benchmark dataset.

GraphCPD: Coherent Point Drift for Point Cloud Registration via Graph Signal Processing

GraphCPD bridges graph signal processing (GSP) and coherent point drift (CPD) probabilistic registration by extracting rigid-transformation-invariant high-frequency components to form 9D graph signals, substituting isotropic GMM covariances with local graph Laplacians, and pruning correspondence search via high-frequency priors to achieve state-of-the-art accuracy with order-of-magnitude speedups on large-scale point clouds.

GRF-Recon: Global Ray-Field Optimization for Long-Sequence Feed-forward Reconstruction

GRF-Recon introduces re-normalized LoRA geometric prior injection and a hybrid-weight sparse ray-field factor graph optimization to enable low-drift, globally consistent kilometer-scale feed-forward 3D reconstruction from uncalibrated monocular videos within a strict 24 GB GPU budget.

HHA: Hierarchical Hyperbolic Constraints for Imperceptible Point Cloud Attacks

HHA decomposes 3D point clouds into dual geometric and semantic hierarchies and embeds them into Poincaré hyperbolic space, applying level-wise hyperbolic Chamfer constraints to preserve multi-scale structure and achieve state-of-the-art imperceptibility while maintaining near-100% attack success rates.

Holo-Captioning: A Comprehensive Textual View of 3D Scenes

This paper introduces Holo-Captioning, a novel task striving for the textual equivalent of 3D scenes, and develops HoloScribe, a decoupled 3D-LLM that jointly predicts semantic categories, 3D bounding boxes, fine-grained attributes, and inter-entity relations in pure text without external detectors.

Identity-Preserving Human Reconstruction from a Single Image via 3D Token Inference

This paper introduces the Identity-Preserving Large Human Reconstruction Model (IPRM), a feed-forward framework that operates on an SMPL-based sparse voxel space to infer invisible regions via visibility-based cross-attention while strictly preserving visible identity tokens, paired with a parallel 3D ID Adapter to eliminate identity drift during decoding, achieving photorealistic 3DGS and mesh reconstruction in approximately 0.6 seconds.

IndoorSplat: Enhanced Indoor Scene Reconstruction with Structured 2D Gaussian Splatting

To tackle extensive textureless surfaces, severe geometry-appearance ambiguity, and sparse SfM initialization failures in indoor scenes, IndoorSplat builds upon 2D Gaussian Splatting by introducing feed-forward dense initialization with locally aligned monocular depth completion, an iterative adaptive densification strategy tracking active primitives with soft L1 opacity regularization, and scale-invariant depth-normal supervision, establishing new state-of-the-art geometry reconstruction benchmarks on ScanNet and ScanNet++.

InSeg: Interactive Refinement via Intent Propagation for Point Cloud Semantic Segmentation

Addressing the "local correction trap" of single-point supervision in point cloud semantic segmentation, InSeg conceptualizes sparse user clicks as regional intent seeds, derives denoised intent-consistent regions via feature affinity, spatial connectivity, and instance priors to generate virtual clicks, and achieves stable online fine-tuning via interaction-aware entropy regularization and history replay.

Instance Segmentation as Tracking: A New Paradigm for Multi-Small-Object Tracking with Event Cameras

To overcome temporal discretization and heavy background redundancy in traditional event-based "convert-then-detect-and-track" paradigms, this paper reformulates multi-small-object tracking as an instance segmentation task on 3D spatio-temporal event point clouds, introducing Ev-ISNet with fine-grained 3D sparse convolutions and geometry-consistent graph clustering for high-accuracy, streaming tracking.

InstantHDR: Single-forward Gaussian Splatting for High Dynamic Range 3D Reconstruction

Addressing the failure of existing feed-forward 3D reconstruction models under multi-exposure low dynamic range (LDR) inputs, InstantHDR proposes the first feed-forward 3D Gaussian Splatting framework that reuses frozen geometry attention to guide irradiance fusion and leverages a meta-network for scene-adaptive tone mapping, initializing HDR 3D scenes in seconds.

InstGS: Shared-Template Gaussian Instancing for Object-Redundancy-Free Rendering

InstGS introduces GPU-accelerated instanced rendering to 3D Gaussian Splatting by extracting a shared geometric template via gradient-driven instance segmentation and applying learnable lightweight offsets in position, base color, and opacity, achieving massive memory savings and higher frame rates with negligible fidelity loss in repetitive scenes.

KineticGS: Momentum-driven Coherent 4D Gaussian Splatting for Monocular Dynamic Scene Reconstruction

Addressing trajectory fragmentation and non-physical deformations caused by the lack of physical priors in monocular dynamic reconstruction, KineticGS models Gaussians as momentum-carrying particles within a hierarchical framework spanning momentum-guided sampling, learnable energy-flow activation, and holistic kinetic synchronization, yielding +0.91 dB PSNR gains across dynamic object regions.

Large-Scale Light Field Synthesis from Videos Enables Geometrically Consistent Bokeh Editing

Addressing the failure of depth-based bokeh editing on complex geometries such as reflections and intricate occlusions, Vid2Bokeh presents a scalable data acquisition pipeline that reconstructs dense 25×25 light fields from casual videos via feed-forward 3D reconstruction to bypass depth estimation entirely, producing a 100K-scene dataset that empowers a diffusion model to perform robust, geometrically consistent bidirectional focus and depth-of-field editing.

Latent Fusion: Decoding Consolidated 3D Geometry from Feed-forward Geometry Transformer Latents

This paper introduces Fus3D, which discards the conventional predict-then-fuse paradigm in feed-forward geometry transformers and instead leverages learned canonical voxel queries with interleaved cross/self-attention to decode dense Signed Distance Fields (SDF) directly from intermediate multi-view latents (VGGT), achieving pose-free, artifact-free, and complete 3D surface reconstruction.

Learning Generatable Mutual Distance for Scene-Aware Human Motion Generation

Addressing the spatial ambiguity of mutual distance in stochastic generation due to the absence of an absolute spatial coordinate frame, this paper proposes MDNet—a two-stage diffusion framework that introduces vision-language model (VLM) semantic spatial anchors to resolve geometric ambiguity and synthesizes smooth mutual distance trajectories in the frequency domain to guide physically plausible, temporally coherent scene-aware human motions.

Learning Geometry-Aware Embedding Fields for Intrinsic Riemannian Mappings

To tackle the challenge of computing intrinsic Riemannian mappings on non-manifold and noisy point sets, this paper presents a data-driven framework that learns a continuous geometric embedding field via sparse octree convolutions and decodes exponential and logarithmic maps in a single forward pass using triplane hyper-modulation.

LESV: Language Embedded Sparse Voxel Fusion for Open-Vocabulary 3D Scene Understanding

Addressing spatial ambiguity and high preprocessing costs in 3D Gaussian Splatting, LESV establishes a deterministic confidence-gated fusion framework using explicit Sparse Voxel Rasterization and dense AM-RADIO foundation model features, slashing preprocessing overhead while achieving SOTA open-vocabulary 3D scene understanding.

LiteGS: a high-performance framework to train 3dgs in subminutes via system and algorithm codesign

LiteGS tackles the systemic bottlenecks of 3DGS training—atomic gradient contention, disordered memory layout, and ambiguous densification metrics—via warp-based rasterization, a Cluster-Cull-Compact memory pipeline, and opacity gradient variance guidance, achieving sub-minute training (47–67s) on an RTX 3090 while matching or surpassing SOTA visual fidelity.

LOOM: Weaving Geometry-Consistent Human-Object Interaction Videos via Progressive Curriculum Learning

ByteLOOM tackles geometric distortion and hand mesh dependency in human-object interaction video synthesis by introducing an RCM-cache mechanism to formulate view generation as spatial texture retrieval, combined with a progressive three-stage curriculum that yields highly consistent 6-DoF interaction videos conditioned only on simplified 2D skeletons.

LumiDepth: Stable Monocular Depth in Multi-Illumination Scenes

Addressing geometric distortion and temporal inconsistency under real-world multi-light and shifting illumination conditions, LumiDepth presents a target-domain label-free framework that learns illumination-invariant monocular depth from multi-illumination RGB image groups via Disagreement-Calibrated Probabilistic Pseudo Supervision (DCPS) and Frequency-aware Consistency and Distillation (FaCD).

LVSPM: Long Sequence View Synthesis and Pose Estimation Model

LVSPM introduces an end-to-end forward sequence prediction framework based on Large-Chunk Test-Time Training (LaCT) and learnable camera tokens, achieving accurate pose estimation and real-time novel view synthesis across uncalibrated sequences up to 256 frames without requiring dense 3D ground truth.

MagnetGS-Mesh: High-Quality Multi-Object Mesh Reconstruction via Adaptive Surface Optimization

Addressing surface holes and outlier scatter caused by object-wise 3DGS decomposition, MagnetGS-Mesh introduces Enhanced Occupancy Learning and Adaptive Local Normal Consistency (ALNC) Loss to magnetically relocate outlier Gaussians into sparse surface areas, enabling high-quality, lightweight, and artifact-free multi-object mesh extraction.

Manifold-Aware Spectral Compaction: A Graph Signal Processing Perspective on Online Gaussian Reduction for 3DGS SLAM

This paper introduces Manifold-Aware Spectral Compaction (MASC), a principled online reduction framework for 3DGS SLAM that reformulates Gaussian reduction as spectral signal reconstruction on a Riemannian manifold via Symmetrized KL topology modeling, incremental Nyström embedding, and closed-form Bures-Wasserstein barycenters, achieving a 2.8× GPU memory reduction, 2.4× rendering speedup, and an ATE of 0.98 cm on TUM RGB-D while discarding ~60% of redundant primitives.

MindBlock: Probing Spatial Assembly and Structure in Unified Multimodal Models

Addressing the limitation of spatial reasoning in unified multimodal models to passive 2D VQA, this paper introduces MindBlock, a benchmark focusing on pixel-space active construction, alongside 3DGS-Eval, a volumetric consistency evaluation protocol using 3D Gaussian Splatting to uncover the 2D-3D consistency gap, spatial drift, and the perception-execution gap.

Minute4D: Training High-Fidelity 4D Gaussian Splatting in One Minute

To tackle camera-object motion ambiguity, view coverage scarcity, and prolonged optimization in casual monocular dynamic 3D reconstruction, Minute4D combines foundation-model-driven segmentation-tracking geometric completion with multi-view photometric loss-guided 4D Gaussian density control, compressing 4D Gaussian Splatting optimization time to ~40 seconds (a 5x+ speedup) while establishing state-of-the-art rendering fidelity and memory efficiency.

MoBa-GS: Learning a Spatially-Varying Motion Basis over a Dynamic Canonical Space for 4D Reconstruction

MoBa-GS resolves 4D dynamic reconstruction bottlenecks by introducing a structural inversion from global temporal trajectories to a spatially-varying motion basis, factorizing non-rigid deformations into a low-frequency dynamic canonical space and sparse linear combinations of local motion primitives, achieving state-of-the-art fidelity from random initialization without SfM point cloud priors alongside real-time rendering (>163 FPS) and 16x faster training via time-invariant caching.

Modeling and Compensating Phase Error in High-speed 3D Reconstruction

The paper formulates an explicit theoretical model revealing that error diffusion dithering in binary defocusing introduces systematic phase errors correlated with diffusion direction but invariant to phase-shifting steps and defocusing levels, and proposes an Error Diffusion Fusion (EDF) strategy pairing forward and inverse Floyd-Steinberg kernels in four-step phase shifting to achieve complete inter-fringe error cancellation at zero runtime overhead.

MotionEditGS: Editing Motion and Appearance of 4D Scenes from Monocular Video via Semantically Anchored Gaussians

Addressing the under-constrained nature of monocular dynamic scene editing where prior image-diffusion methods collapse into shallow appearance-only modifications, MotionEditGS integrates video diffusion priors with semantically anchored 3D Gaussians and gradient-driven feature lifting, enabling photorealistic, temporally coherent 4D motion and appearance edits.

MotionSplicer: Part-Based Motion Editing for 4D Volumetric Videos

Addressing the inability to interactively manipulate motions in 4D Gaussian volumetric videos with multiple objects and non-rigid parts, MotionSplicer introduces an annotation-free, template-free part-based editing framework that discovers 3D motion parts via temporal difference prompts and adaptive 3D unify-and-split, while learning spatiotemporally coherent Gaussian skinning weights through a time-invariant neural coordinate feature grid and temporal consistency refinement.

MuCHeR: Multi-Person Camera-Centric Human Detection, Mesh Recovery and Tracking

MuCHeR (Multi-HMR 2) is a simple and robust DETR-based one-stage framework for multi-person 3D human detection, mesh recovery, and tracking that significantly reduces training cost and enables heterogeneous 2D supervision via 2D-location Hungarian matching, achieves precise camera-centric metric 3D localization without ground-truth intrinsics through predicted vertical FOV, and achieves online multi-person tracking without any video training by distilling SAM2 memory encoder features.

Multi-view Multi-vehicle Driving Dataset for Novel View Synthesis

Addressing the fundamental limitation that existing driving NVS benchmarks evaluate only temporal interpolation along a single vehicle trajectory, this paper presents MV2, the first real-world multi-vehicle driving dataset captured synchronously from a four-wheeler, a two-wheeler, and an aerial drone, introducing a rigorous cross-trajectory and aerial-to-ground extrapolation benchmark where state-of-the-art models experience dramatic performance degradation under wider baselines.

NanoGS: Training-Free and Lightweight Gaussian Splat Simplification

NanoGS introduces a training-free and lightweight 3D Gaussian Splatting simplification framework that constructs a spatial sparse KNN merge graph, scores candidate merges via an information-theoretic I-divergence and appearance metric, and fuses primitives using closed-form Mass-Preserved Moment Matching (MPMM), achieving extreme primitive compaction up to 1000x without requiring multi-view images or GPU retraining.

NeLU3D: Neural Inverse Structured Light without Modeling the Projector

NeLU3D presents a neural inverse structured light reconstruction framework that operates without explicit geometric or radiometric modeling of the projector, baking a one-time 3D-to-color mapping into a multiresolution hash-encoded MLP and optimizing a signed distance function with physically bounded ambient decoupling to deliver sub-millimeter surface reconstructions from as few as three suboptimal patterns and a white flash.

NEOMAP: Novel-View Synthesis via Noise Initialization by Manifold Alternating Projection

To tackle view misalignment and generation artifacts in single-image and monocular video novel-view synthesis, NeoMap discovers high-fidelity view-consistent solutions directly within the native video data manifold of pre-trained models by optimizing the initial Gaussian noise via manifold alternating projection iterations without task-specific training or online guidance.

NeuralGarSim: Geometry-agnostic Garment Simulation with Neural Fields

Addressing the mesh-resolution sensitivity of conventional solvers and the strict 2D parameterisation barrier in existing neural simulators, NeuralGarSim reformulates nonlinear Kirchhoff–Love thin-shell mechanics directly in 3D Euclidean space via Tangential Differential Calculus (TDC), achieving continuous, parameterisation-free quasistatic garment simulation seamlessly across multi-panel topologies and diverse geometric representations including meshes, UDFs, point clouds, and 3D Gaussians.

Nexels: Neurally-Textured Surfels for Real-Time Novel View Synthesis with Sparse Primitives

Addressing the primitive explosion of 3D Gaussian Splatting when representing high-frequency appearance, Nexels decouples geometry and appearance by modeling surfaces with generalized quad-indicator surfels and querying an Instant-NGP neural field on only the top-\(K\) contributing fragments per pixel, delivering 50 FPS real-time rendering and state-of-the-art perceptual fidelity with up to \(31\times\) fewer primitives.

Novel View Synthesis as Video Completion

FrameCrafter casts sparse novel view synthesis as low-frame-rate video completion, converting pretrained video diffusion models into strictly permutation-invariant multi-view generators via per-view independent VAE encoding, query-centered Plücker ray conditioning, and temporal RoPE removal—outperforming image diffusion baselines trained on massive multi-view datasets using only 1K scenes for fine-tuning.

ObjectForesight: Predicting 3D Object Trajectories from Human Videos

ObjectForesight is an object-centric 3D forward dynamics model that predicts physically consistent future 6-DoF object trajectories from passive egocentric videos, powered by an automated curation pipeline yielding 2+ million clean trajectories and a geometry-conditioned Diffusion Transformer.

OCA: ODE-Driven Cross-Attention for Image-to-Point-Cloud Registration

Addressing the fundamental attention ambiguity caused by cross-modal discrepancy in image-to-point-cloud (I2P) registration, this paper formulates cross-attention as continuous assignment ordinary differential equations (ODEs) and introduces a non-parametric, plug-and-play module (OCA) that iteratively purifies 2D-3D correspondences via dynamical propagation with negligible runtime overhead (+6ms).

OmniLife360: A Benchmark for 3D Reconstruction from In-the-Wild 360° Captures

To tackle severe camera ego-motion and dynamic interference in long-horizon captures from consumer 360° cameras, this paper presents OmniLife360—a benchmark of 2,407 4K minute-level in-the-wild panoramic sequences—along with Omni4DGS, the first spherical 4D Gaussian Splatting framework that decomposes radial/tangential velocities and decouples dynamic foreground via mask supervision.

OmniRen: Neural Rendering wih Heterogeneous Scene Primitives

RenderFormer-V2 (OmniRen) addresses the scalability and primitive diversity limitations of neural rendering by combining Hilbert space-filling curve sliding window attention with physics-informed attention sinks and a BRDF-agnostic neural material appearance space, enabling feed-forward global illumination rendering across heterogeneous primitives (>100k) without per-scene training.

One-Shot Feed-Forward 360° Animatable Avatar via Inpainted UV-Space Gaussian Modeling

FHAvatar is a one-shot feed-forward 360° animatable full-head avatar creation framework that leverages 3D GAN priors and symmetric UV-space inpainting to regress rigged 3D Gaussians on parametric meshes, achieving high-fidelity 360° rendering and real-time animation at 246 FPS.

Online 3D Instance Segmentation at task-oriented granularity with Unposed Monocular Video

Addressing the online perception needs of embodied agents on unposed monocular video, this paper introduces a task-oriented 3D instance segmentation system that reuses feed-forward dense pixel correspondences from MASt3R-SLAM to build geometric association scores and intra-frame mutual exclusivity constraints, achieving real-time 3D instance disentanglement via confidence-priority clustering without requiring depth sensors or offline pose priors.

Optimizing Mesh Animation from Video via Shape Flow Guidance

Addressing the severe shape collapses and motion artifacts caused by conventional 2D rendering-based reconstruction losses in video-driven mesh animation, this paper proposes Shape Flow Guidance (SFG) to extract temporally coherent 3D shape sequences via training-free intervention on a pretrained mesh generator's sampling trajectory, paired with a decoupled skeletal animation model for high-fidelity 4D mesh optimization.

ORACLE-3D: Open-world Region-aligned Cross-modal Learning for Label-efficient 3D Scene Understanding

ORACLE-3D presents a unified multi-modal contrastive learning framework that uses images as a semantic intermediary, employing dual-level cross-modal distillation and a Vision-Point Matching module to dynamically correct 2D-to-3D projection noise without requiring costly point-text pairs, substantially outperforming prior art in both base-annotated and annotation-free 3D segmentation settings.

OREO: Fidelity Alignment in 3D Generation via On-the-fly Rendering-Editing Optimization

OREO establishes a closed-loop "Render-Edit-Optimize" framework that dynamically generates structure-preserving, high-fidelity 2D pseudo ground truths via reinforced editing and distills visual fidelity gains into 3D feed-forward generators via latent contrastive optimization without gradient attenuation through differentiable renderers.

PaD-GS: Leveraging Distortion Map for Panoramic Gaussian Splatting

PaD-GS introduces an analytical distortion map derived from panoramic imaging geometry and incorporates it into Gaussian learning via a whole-to-partial paradigm (distortion-aware decoding and adaptive opacity modulation), significantly eliminating rendering blur and distortion artifacts in 360-degree novel view synthesis.

Parametric SDF for Dynamic Surface Reconstruction

By reformulating static Signed Distance Functions into time-dependent parametric trajectories parameterized by hybrid polynomial-Fourier bases, p-SDF couples differentiable iso-surfacing with physically based differentiable rendering to achieve robust, high-fidelity dynamic surface reconstruction and material disentanglement under large non-rigid deformations and topological changes.

PDF-Omni: Poincaré Dual Disk Distortion Field-based Recurrent Update for Omnidirectional Stereo Matching

To tackle matching failure caused by severe peripheral fisheye distortion and view-transition ambiguity in multi-fisheye seam regions, PDF-Omni introduces a continuous Poincaré Dual Disk distortion field to drive geometry-aware receptive field allocation in recurrent GRU updates, paired with a local entropy cumulative loss to supervise multi-modal cost distributions.

Penetration-Free Compositional 3D Generation via Gaussian Surface Offset

To tackle severe geometric collisions and inter-object penetrations in compositional text-to-3D generation, CompSAG introduces a Gaussian Surface Offset mechanism that computes surface-normal repulsive forces via continuous proxy surfaces to shift penetrating 3D Gaussians, combined with relation-aware score distillation and normal rectification for penetration-free, high-fidelity multi-object scenes.

PhysConvex: Physics-Informed Dynamic Convex Fields for Reconstruction and Simulation

PhysConvex establishes a unified 4D representation by treating dynamic convex hulls as differentiable rendering elements, deformation carriers, and continuum physical supports simultaneously, coupling boundary-driven kinematics with mesh-free neural skinning reduced-order simulation for highly efficient dynamic video reconstruction and system identification.

Physically Grounded Dual-Opacity Gaussian Splatting for Joint RGB-TIR Reconstruction

RaViGS establishes a physically grounded 3D Gaussian Splatting framework for joint RGB-TIR scene reconstruction by parameterizing thermal response through intrinsic physical attributes (emissivity, temperature, reflectance) and decoupling cross-spectral visibilities via dual-opacity rendering, resolving fundamental radiative mismatches and occlusion conflicts on a shared geometric scaffold.

Point Diffusion Mamba: Unified Diffusion-State-Space Modeling for Single-View 3D Reconstruction under Data Scarcity

To tackle ill-posed single-view 3D reconstruction and memory bottlenecks under data scarcity, this paper proposes Point Diffusion Mamba (PDM), unifying local geometric aggregation and dual-directional Mamba diffusion with a hierarchical feature integration network and dynamic weighted sampling, achieving state-of-the-art geometry recovery with only 0.39 GB GPU memory.

PointGT: Simultaneous Geometric and Textural Editing for Point-Based Representations

PointGT combines proximity-attention point rendering (PAPR) with normal-free neural UV parameterization, utilizing geometric regularizers and a deformation-aware canonical correspondence mechanism to achieve simultaneous, decoupled editing of geometry and high-resolution appearance where textures remain firmly attached under non-rigid deformations.

PoseShield: Neural Collision Fields for Human Self-Collision Resolution

PoseShield learns a neural collision field with Eikonal regularization (approximating pose-space SDF) within the SMPL pose space, embedding it as a differentiable constraint function in gradient-based constrained optimizers like SLSQP. This achieves post-processing correction from self-penetrating poses to the nearest collision-free poses. This method scales to motion sequence collision resolution without retraining downstream motion generation models, achieving a 95.8% success rate on the self-built HwC dataset, significantly outperforming COAP (44.6%).

PrimitiveUDF: Primitive-Based Unsigned Distance Fields for Surface Reconstruction from Point Clouds

To resolve the fundamental failure mode of unsigned distance fields conflating nearby layers in multi-surface regions, PrimitiveUDF reorganizes Euclidean neighborhoods into 2D local charts via viewpoint-guided spherical projection and constructs hybrid triangle/segment primitives as an explicit geometric bias for distance regression.

PriSplat: Propagating Reliable Multi-view Information for Distractor-Free 3DGS

PriSplat moves beyond the naive zero-masking paradigm in distractor-free 3DGS by repurposing a large-scale novel view synthesis prior into a 3D-aware inpainting harmonizer guided by geometry-aware support view selection and mask-aware fast-weight updates, achieving a state-of-the-art mean PSNR of 23.11 dB on the On-the-go benchmark while effectively eliminating floaters.

Probing the 3D Object-Level Understanding of Pre-Trained Detection Transformers

This paper introduces a probing framework to investigate the implicit geometric representations of 2D detection transformers, revealing that DETR query embeddings spontaneously encode accurate object-level depth and camera-frame 3D coordinates without any 3D supervision during pre-training.

Progressive and Localized Super-Resolution of 3D Objects via Localized Latent Voxel Diffusion

Addressing the cubic memory and compute bottleneck of full-object 3D diffusion generation, PLSR introduces a progressive and localized super-resolution framework on structured sparse voxel latents, decomposing global refinement into aligned patch triplets, lightweight low-resolution cross-attention LoRA fine-tuning, and per-step iterative denoising stitching to super-resolve coarse meshes up to \(2560^3\) resolution on a single GPU.

P²Fusion: Prompt-based Progressive Infrared-Visible Image Fusion via Dual-Prior Distillation

To tackle optimization conflicts from static prior penalties and granularity mismatch from foundation models in infrared-visible image fusion, P²Fusion introduces the "Teach-to-Fuse" paradigm that distills intrinsic thermal saliency and spatial quality into dynamic learnable prompts, combining with a Gated Dynamic Expert Recalibration module for decoupled feature refinement.

R3RECON: Radiance-Field-Free Active Reconstruction via Renderability

Addressing the high computational latency and early-stage instability of radiance-field-dependent active mapping, R3CON introduces a closed-form renderability metric built upon a lightweight voxel-statistics map and 360° panoramic utility aggregation, achieving millisecond-level NBV scoring and superior 3DGS novel-view quality under matched budgets with sub-300MB GPU memory.

RaPTGS: Render-Agnostic Post-Training Compression of 3D Gaussian Splatting

Under a strict model-only constraint where training images, camera poses, and rendering pipelines are unavailable, RaPTGS introduces multi-criteria render-agnostic pruning (opacity, shape anisotropy, scale-aware appearance energy), training-free voxel mass compensation, degree-wise SH vector quantization, and spatial-reordering entropy coding to deliver a 30x compression ratio in seconds while preserving high rendering fidelity.

Raymap-Guided Coupling for Drift-Robust Unposed Feed-Forward 3D Reconstruction

To resolve the critical problem of cumulative camera pose drift that causes geometric collapse and blurry rendering in long-sequence unposed feed-forward 3D Gaussian Splatting, NoDrift3R explicitly anchors 3D Gaussian centers to dense raymap-and-depth geometry, establishes a bidirectional synergistic "Rendering-to-Geometry Gain" loop, and integrates a dual-frequency viewpoint scheduler to achieve drift-robust, globally consistent 3D reconstruction across extended trajectories.

RealDyadic: Synthesizing Realistic Dyadic 3D Dialogue with Neural Appearance Priors

Addressing the severe lip misalignment and mouth closure failures in dyadic 3D avatar generation caused by noisy 3D tracking pseudo-labels, RealDyadic introduces a two-stage framework coupling diffusion-based motion synthesis with 3D Gaussian Splatting rendering, leveraging 2D photometric gradients during alternating training to correct underlying 3D motion priors beyond tracking ceilings.

Reconstructing 3D Human-Object Interaction via a Unified Triplane Space

To tackle the structural disparity between parametric human templates and diverse object geometries in monocular 3D human-object interaction (HOI) reconstruction, this paper introduces a unified triplane feature space, coupling HOIT-VAE implicit priors, camera-aware HOIT-T image-to-triplane prediction, and SMPL-H pose-fitting to deliver balanced, highly accurate interaction meshes.

REFINE: Super-efficient Pruning for 3D Gaussian Splatting via Rendering-Free Primitive Importance

Addressing the trade-off between inaccurate parameter-space heuristics and computationally prohibitive render-aware scoring, REFINE presents a completely rendering-free primitive importance metric by analytically decoupling the Hessian field into closed-form view-dependent visibility and geometric projection energy, achieving a >3,000x reduction in pruning FLOPs with SOTA rendering fidelity.

ReGen3D: Generalizable Unified Representation Learning for 3D Understanding

Addressing the vulnerability of existing 3D models tied to a single geometric format under representation shifts, this paper formalizes the new setting of Representation Generalization (RG) and presents ReGen3D, a unified framework integrating Unified Representation Tokenization (URT), Shared Memory Interaction (SMI), and Efficient Representation Adapter (ERA) to achieve cross-representation pretraining, decoupled single-format test-time inference, and lightweight adaptation to unseen geometric primitives.

RegVGGT: Sustainable Visual Geometry Grounding for Streaming via Regulated Memory

RegVGGT proposes a training-free token regulation framework for streaming feed-forward 3D reconstruction, exploiting the temporal consistency of token saliency to permanently retain first-frame global anchors, decouple transformer layers, admit at most 1% salient tokens per incoming frame, and recover exact attention scores via FlashAttention-compatible LSE statistics, enabling sustainable long-horizon 3D grounding on consumer-grade GPUs.

ReInGS: Re-Initializing 3D Gaussians against Sparsity Discrepancy in Few-Shot Novel View Synthesis

ReInGS tackles the overlooked sparsity discrepancy and data leakage problem in few-shot 3DGS caused by dense SfM points from extended views, proposing a two-stage coarse-to-fine Gaussian re-initialization framework combining expansion-based hybrid point sampling and view-dominant details-aware sampling to achieve state-of-the-art novel view synthesis purely from sparse training views.

REON-NVS: Real-Time Online Novel-View Synthesis from Sparse-View Videos

Addressing the computational bottleneck, reliance on iterative camera pose optimization, and requirement for dense camera arrays in dynamic streaming NVS, REON-NVS introduces the first fully feedforward online framework that couples self-supervised latent pose estimation and mapping with a Bi-SSM-based temporal scene reconstructor, rendering photorealistic novel views at 32 FPS on a single A100 GPU without accumulated delay.

RePer-360: Releasing Perspective Priors for 360\(^\circ\) Depth Estimation via Self-Modulation

RePer-360 proposes a self-conditioned framework based on normalization modulation. Without overwriting pre-trained perspective priors, it utilizes complementary projection (CP and ERP) features as guidance signals for geometric alignment to modulate a depth foundation model, achieving state-of-the-art (SOTA) omnidirectional depth estimation with an extremely small amount of panoramic data (approx. 1%).

ReSWD: ReSTIR‘d, not shaken. Combining Reservoir Sampling and Sliced Wasserstein Distance for Variance Reduction

Inspired by the ReSTIR resampling algorithm in real-time computer graphics path tracing, this paper integrates Weighted Reservoir Sampling (WRS) into the Sliced Wasserstein Distance (SWD) framework, adaptively preserving highly informative projection directions across optimization steps to boost gradient signal-to-noise ratio and accelerate convergence while strictly maintaining Monte Carlo unbiasedness.

RoomPlanner: Reachability-Aware View Sampling for Text-to-Room 3D Gaussian Splatting

Addressing the mismatch between spatial layout feasibility and camera view selection in text-to-room generation, RoomPlanner introduces hierarchical LLM planning with dual collision-and-reachability constraints, seamlessly bridging traversable layout maps to ReachView camera sampling and Interval Timestep Flow Sampling (ITFS) to synthesize physically plausible, decoupled, and editable 3D Gaussian Splatting indoor scenes in under 30 minutes from short text.

Rosetum3D: A Large-Scale 3D Vision Dataset from Preharvest Roses

Addressing the acute lack of 3D benchmarks in floriculture automation, this paper introduces Rosetum3D, comprising 21,114 high-resolution RGB-D images and 46,848 annotated rose instances across commercial greenhouses, establishing a botanically guided 2D-to-3D backprojection labeling paradigm and multi-task benchmarks spanning detection, pose estimation, and temporal 3D localization.

SAND: Stage-Aware Noise Decomposition for Training-Free Diffusion Guidance

Addressing the flaw of static attention perturbations in training-free guidance that overlook diffusion denoising dynamics, SAND orthogonally decomposes Gaussian noise into magnitude and directional components and modulates them with a time-adaptive schedule, significantly enhancing global structural coherence and fine-grained geometric details across image and video models without retraining.

Scale3D: Autoregressive Modeling for Large Outdoor Scene Generation

Scale3D decouples global layout planning from local geometric synthesis via a hierarchical autoregressive framework, combining an enhanced high-fidelity 3D chunk VAE and Direct Preference Optimization (DPO) to synthesize seamless, large-scale outdoor 3D meshes with natural spatial text control.

SEAR: Simple and Efficient Adaptation of Visual Geometric Transformers for RGB+Thermal 3D Reconstruction

Addressing the failure of feed-forward visual geometric transformers (such as VGGT) to maintain cross-modal alignment when jointly processing RGB and thermal imagery, SEAR introduces a lightweight adaptation strategy updating fewer than 5% of parameters via modality-specific camera tokens, alternating-attention LoRA adapters, and unpaired cross-view batch sampling, achieving SOTA multimodal 3D reconstruction and pose estimation with negligible inference overhead.

SegFly: A 2D-3D-2D Paradigm for Aerial RGB-Thermal Semantic Segmentation at Scale

SegFly introduces a scalable geometry-driven 2D-3D-2D paradigm for aerial perception that annotates less than 3% of RGB images, lifts them into a metric 3D point cloud, and renders dense pseudo-labels across 97% of RGB and 100% of thermal images with 91% and 88% accuracy, while simultaneously achieving 87.05% pixel-level RGB-T registration without hardware-level time synchronization on off-the-shelf UAVs.

SFD-Net: Sharp Feature Detection Network Based on Local Geometric Features

SFD-Net proposes a plug-and-play multi-scale Local Geometric Descriptor (LGD) derived from second moments of normal differences, combining an enhanced PointNet++ backbone and an imbalance-aware Focal Tversky-Dice loss to achieve state-of-the-art sharp feature detection under joint density variation and noise.

SharpGS: Sharpness-Preserving 3D Gaussian Splatting with Differentiable Blur-Driven Density Control

Addressing the insensitivity of standard 3D Gaussian Splatting (3DGS) to intricate boundaries and high-frequency textures caused by gradient-only density control, SharpGS introduces blur as an explicit perceptual cue via a CUDA-accelerated differentiable blur framework, substantially elevating fine rendering quality while keeping the primitive representation compact through dual regularizations.

SkySplat-OV: Generalizable Language Gaussian Splatting for Open-Vocabulary Scene Understanding from Sparse Satellite Views

The first self-supervised, generalizable language 3DGS framework designed for sparse satellite views that incorporates rational polynomial coefficient (RPC) camera models into a feed-forward pipeline, leverages Geometry–Radiance–Semantics Fusion (GRS-Fusion) alongside relative height supervision, and employs single-stage SAM3 distillation to achieve up to 1070× inference speedup over per-scene optimization while delivering state-of-the-art open-vocabulary 3D semantic segmentation.

SMP-UWGS: Coupled Physics-Geometry Optimization for Scalable Multi-Partition Underwater 3D Reconstruction

To tackle computational explosion and color degradation caused by the tight coupling of light attenuation and depth in large-scale marine environments, SMP-UWGS couples a scalable multi-partition Gaussian architecture with a differentiable physical rendering network (DPR-Net), enabling high-fidelity, scalable, and real-time underwater 3D reconstruction.

Steering 3D Generations: Preference Alignment via Direct Reward and Preference Optimization

Addressing severe geometric artifacts and intractable human preference formalization in single-image-to-3D generation, this paper proposes a preference alignment framework combining Gaussian-curvature neural SDF data curation with Direct Reward Optimization (DRO) and Direct Preference Optimization (DPO) on rectified flow, achieving high fidelity and controllable geometry via parameter-efficient LoRA.

StereoEdit: A Diffusion-Based Framework for Stereo-Consistent Image Editing

Addressing cross-view texture drift and geometric distortions in independent 2D diffusion edits, StereoEdit introduces a training-free framework that incorporates Geometrically-Gated Attention Reference (GGAR) at the feature level and Geometry-Aware Latent Alignment (GALA) at the latent level to enforce rigid stereo consistency without retraining.

StratoSplat: Taming Layered Regularities for Sparse Aerial 3D Gaussian Splatting

StratoSplat exploits the inherent gravity-aligned stratification of aerial scenes, coupling layered depth image-guided test-time diffusion adaptation for dense point initialization with a hybrid neural multi-plane Gaussian representation to suppress view-aligned degeneracies under sparse drone captures.

Structure Gaussian Splatting SLAM

Addressing pose drift and floater artifacts caused by vanishing photometric gradients in low-texture and repetitive indoor scenes, this paper explicitly models ubiquitous planes as persistent Planar Gaussian Instances (PGIs) and introduces a "structure closure" mechanism estimating relative camera poses from non-parallel planes without trajectory revisits, achieving sub-centimeter tracking accuracy and photorealistic dense reconstruction.

SupIR-GS: Thermal Infrared Super-Resolution Novel View Synthesis with Imaging-Calibrated 3D Gaussian Splatting

Addressing the core challenge of "signal-to-noise ratio inversion" in thermal infrared imaging—where thermal diffusion smooths true boundaries while sensor non-uniformity introduces structured noise—SupIR-GS presents the first imaging-calibrated super-resolution 3D Gaussian Splatting framework, decoupling hardware artifacts via a differentiable physical calibration layer, boosting flat thermal gradients via a local tone adapter, and progressively recovering fine geometric boundaries via frequency-aware curriculum learning.

SWSL: Semantic-aware Weakly Supervised Learning for 3D Motion Generation using 2D Motion Data

To alleviate the severe scarcity of paired text-3D motion data, SWSL constructs a unified text-2D-3D semantic embedding space, refines coarse 3D pseudo-labels with 2D motion embeddings, and directly optimizes end-to-end motion generators via reinforcement learning using semantic discrepancy rewards, cutting 3D data dependency while markedly enhancing motion fidelity and text-motion alignment.

SyncFix: Multi-View Consistent Diffusion Refinement of 3D Reconstructions

To eliminate cross-view geometric and semantic inconsistencies caused by independent 2D diffusion refinement in sparse-view 3D reconstruction, SyncFix reformulates multi-view repair as a joint latent bridge matching problem, using permutation-invariant cross-view attention to achieve pose-free synchronization and single-step deterministic inference with superior 3D consistency.

SynLF: Zero-Shot Metric Depth from Light Field Cameras via Physics-Grounded Synthesis

To overcome data scarcity and poor real-world robustness in light field depth estimation, SynLF synthesizes supervisory light fields on-the-fly from large-scale RGB-D data using a physics-grounded pipeline (PG-LF) modeling defocus and non-Lambertian perturbations, coupled with VisDepth, an iterative estimator with explicit visibility reasoning that achieves zero-shot metric depth (MAE 30.05 mm, AbsRel 1.38%) on real captures without fine-tuning.

Tackling Misattribution in 3D Intrinsic Decomposition via Proximity Attention Point Rendering

Reveals the root cause of the misattribution issue in point-based intrinsic decomposition due to volume rendering aggregation over translucent primitives, and proposes Intrinsic PAPR, which leverages Proximity Attention Point Rendering for direct per-point supervision along with cIMLE ambiguity-aware priors and space carving loss to achieve multi-view consistent, high-fidelity, and editable 3D representations.

Taming LLMs for Codematic Indoor Scene Generation

Addressing the core issue where fine-tuning LLMs for indoor 3D scene code generation leads to prompt neglect caused by token-heavy layout history, SceneSpinner curates the first large-scale open-set dataset of 52K rooms and introduces a language-based planning stage alongside a Conditional Mutual Information (CMI) regularization objective, substantially improving fine-grained layout precision and instruction-following fidelity.

Tempo-SAM3D: Monocular Video to 4D via Temporal Memory-Guided Generation

Tempo-SAM3D introduces the first training-free framework that lifts layout-aware single-image 3D generation to temporally coherent 4D generation from monocular video by synergizing dual-path temporal memory caches, solver-level velocity correction, and global HexPlane temporal smoothing.

The Sterkfontein Caves Dataset: A Novel View Rendering Challenge from the Cradle of Humankind

Focusing on the extreme low light, high dynamic range, and intricate karst rock textures of the UNESCO World Heritage Sterkfontein Caves, this paper presents the first multi-view RAW benchmark and reveals that an adapted Raw-Nerfacto baseline reliably outperforms existing specialized low-light and 3D Gaussian Splatting pipelines.

ThermoGS: Decoupling Physical Surface Attributes for Spatio-Temporal Thermal Field Emulation via 4D Gaussian Splatting

ThermoGS integrates 4D Gaussian Splatting with thermodynamic energy balance laws and PINNs to decouple intrinsic surface properties (emissivity, convective heat transfer coefficient, volumetric heat capacity, and absorptivity), achieving physically-consistent all-weather spatio-temporal thermal field emulation with temperature prediction error under 1°C.

TopoFuse: Topology-Aware Tri-Planar Fusion for 3D Cryo-Electron Tomography Segmentation

Addressing fragmented membranes and collapsed organelles caused by ultra-low SNR and missing-wedge artifacts in 3D cryo-electron tomography, TopoFuse reframes topology enforcement from soft-loss penalties into a forward-pass differentiable projection operator coupled with tri-planar SAM encoders and FiLM fusion, achieving a 54% reduction in Betti Number Error while editing only 3.1% of voxels.

TopoGAT: Plug-and-Play Topological Graph Attention for Fine-Grained 3D Segmentation

To resolve ambiguous part boundaries caused by locally self-similar geometry in fine-grained 3D point cloud segmentation, TopoGAT vectorizes persistent homology features and modulates graph attention weights via FiLM, delivering up to a 1.0% mIOU gain on ShapeNetPart and 3.3% on S3DIS with only 1.4M additional parameters.

UniGeo: Unifying Geometric Constraints for Camera-Controllable Image Editing via Video Priors

Addressing structural collapse and cross-view drift caused by fragmented geometric guidance in camera-controllable view synthesis, UniGeo systematically integrates unified geometric guidance across representation, architecture, and loss function levels within a video diffusion foundation model, achieving state-of-the-art geometric consistency under continuous extensive camera trajectories.

VideoSfM: Exploiting Temporal Structure for Video-Based Structure-from-Motion

VidMap (VideoSfM) combines the sequential tracking constraints of SLAM with the non-causal global optimization of offline SfM, achieving drift-free metric 3D reconstruction on long, unconstrained, and uncalibrated videos through provenance-aware data association and metric depth regularization.

ViewSplat: View-Adaptive 3D Gaussian Splatting for Feed-Forward Synthesis

ViewSplat introduces a "view-adaptation" step to pose-free feed-forward 3DGS. It first predicts a set of base Gaussians and scene-conditioned View MLPs. During rendering, these MLPs take target view coordinates to dynamically calculate residual offsets for all Gaussian attributes (position, scale, rotation, opacity, color) on the fly. While maintaining real-time rendering (the largest backbone still achieves 90 FPS), this approach reconstructs high-frequency view-dependent effects such as specular highlights and sharp reflections that are typically lost in static representations, achieving a new state-of-the-art in pose-free NVS fidelity.

VisTa3D: A Dataset and Benchmark for Vision, Tactile, and 3D Point Clouds-based Thin Object Reconstruction

The authors introduce VisTa3D, the first multimodal benchmark for thin-object 3D reconstruction comprising synchronized RGB images, depth maps, tactile response maps, and high-fidelity structured-light ground truth, benchmarking 11 modern reconstruction models and establishing the first visual-range-tactile baseline, Tactile-DC.

VKSR: Scalable Kernel Surface Reconstruction Using Vecchia's Approximation

Addressing the excessive computational cost and oversmoothing of traditional kernel surface reconstruction on dense or large-scale point clouds caused by global low-rank approximations, VKSR introduces Vecchia's approximation from Gaussian Processes to shift computation from a global solve to fully localized response-process inference, reconstructing 14M+ points in minutes with up to 180× speedup over existing kernel baselines.

VoxAnchor: Explicit Voxel-Semantic Grounding for Spatial Understanding in Videos

To tackle spatial conflation and the lack of absolute metric scale caused by 2D projective collapse in Video MLLMs, VoxAnchor explicitly unprojects visual tokens and Depth Anything 3 (DA3) metric priors into a unified 3D voxel grid and constrains feature routing via Voxel-Aware Attention, establishing new SOTA results on VSI-Bench (66.8%) and VSTI-Bench (66.3%) with a compact 4B architecture.

Vulnerability of Privacy-Preserving Visual Localization against Diffusion-based Attacks

This paper introduces a unified privacy attack framework based on multi-scale conditional latent diffusion models with multi-modal LoRA modulation and establishes comprehensive privacy metrics including reconstruction diversity and VLM semantic recovery, systematically demonstrating that geometric and descriptor obfuscation methods remain highly vulnerable to feature inversion while coarse semantic segmentations offer significantly superior privacy preservation.

Wat3R: Underwater 3D Geometry Learning without Underwater Annotations

Wat3R is the first semi-supervised feed-forward 3D geometry estimation framework for complex underwater environments, which synergizes physics-based synthetic underwater rendering with large-scale unlabeled real underwater video mining alongside a static-mask-guided cross-view geometric consistency loss, achieving high-accuracy feed-forward multi-view depth, pose, and point cloud reconstruction without requiring any real underwater 3D annotations.

WildCity: A Real-World Dataset for City-Scale Rendering and Beyond

WildCity organizes 18 continuous driving logs from six U.S. cities into 1,507.1 km of surround-view RGB–LiDAR data and introduces a 3DGS baseline combining rig-aware pose optimization, separate sky modeling, and ground regularization, showing that city reconstruction requires not only room for more Gaussians but also control of pose drift, geometric ambiguity, and off-trajectory rendering errors.

X-SG2S: Safe and Generalizable Gaussian Splatting with X-dimensional Watermarks

X-SG2S splits binary, image, and 3D-object watermarks into patches, uses content-dependent point selection and modality-specific MLPs to write them into an existing 3DGS asset's higher-order spherical harmonic coefficients, and learns to locate and recover the patches without per-scene fine-tuning; the paper reports approximately 3-second injection and 100% unperturbed extraction accuracy in its 48-bit baseline comparison.

Zero-Shot Depth from Defocus: Zero-Shot Metric Depth Estimation from Focus Stacks

This paper proposes FOSSA, a Transformer-based zero-shot depth from defocus (DfD) estimation network. By efficiently aggregating defocus cues in focus stacks through a novel stack attention layer and training on large-scale synthetic data, it achieves strong zero-shot generalization across datasets and scenes. Additionally, the authors release ZEDD—a high-quality real-world depth from defocus benchmark featuring 100 scenes, 4K resolution, and LiDAR ground truth.

Zero-Shot Novel Depth Synthesis Using Foundation Models Scene Representations

Z3D generates target-view representations in the scene-token space of frozen VGGT/WorldMirror, conditioned on source images and target camera poses, then decodes them with a frozen depth head; against the authors' pixel-space depth-diffusion baselines it is more accurate in-domain and on DTU-like settings, but not uniformly better on NRGBD.

∂DIBR: Differentiable Depth Image-based Rendering for Fast Novel View Synthesis

Reconstructs classical Depth Image-Based Rendering (DIBR) into an end-to-end differentiable framework using connected triangle meshes to preserve geometric consistency and topology, reaching 277 FPS on a consumer GPU.