ECCV2026 ECCV2026 accepted papers ECCV2026 paper list AI paper notes top conference papers 3D Vision Image Generation Autonomous Driving LLM Agent Multimodal VLM Robotics & Embodied AI Video Understanding Human Understanding

🎞️ ECCV2026 Accepted Papers¶

452 ECCV2026 paper notes covering 3D Vision (85), Image Generation (64), Autonomous Driving (52), LLM Agent (23), Multimodal VLM (21), Robotics & Embodied AI (20), Video Understanding (20), Human Understanding (19) and other 30 areas. Each note has TL;DR, motivation, method, experiments, highlights, and limitations — 5-minute reads of core ideas.

🦾 LLM Agent (23)¶

Are GUI Agents Focused Enough? Automated Distraction via Semantic-level UI Element Injection: The paper distracts GUI agents with real, content-harmless UI icons subject to spatial and semantic constraints, reaching a five-depth miss rate of 32.99% on UI-TARS-1.5-7B through cumulative multi-candidate search and using injected-icon click statistics to expose grounding weaknesses that content safety alignment alone does not address.
ARGOS: Who, Where, and When in Agentic Multi-Camera Person Search: ARGOS turns person search into budgeted evidence gathering: a four-module agent asks questions and invokes camera-topology tools to isolate a target, with best spatial and temporal Turn-Weighted Success scores of 0.383 and 0.590 across a benchmark of 2,691 tasks.
CiQi-Agent: Aligning Vision, Tools and Aesthetics in Multimodal Agent for Cultural Reasoning on Chinese Porcelains: CiQi-Agent integrates porcelain expertise, local image zoom, and image/text retrieval through two phases of SFT and GRPO, bringing Qwen2.5-VL-7B to 81.5% average multiple-choice accuracy and a 66.7% free-form attribute score on CiQi-Bench; these results measure connoisseurship attributes, not authenticity detection.
Dotting the Eye: An Intent-Driven Image Retouching Agent for Visual Focus Enhancement: EyeControl interprets clicks or coarse strokes as coordinated global and local retouching plans, then trains a diffusion executor with pseudo-intent attention alignment and operation consistency to emphasize the intended region, achieving 21.8845 dB PSNR and an overall score of 8.7054 on ControlArt-Bench without simply maximizing editing strength.
Exploratory, Communicative, and Deployable: Vision-Driven Embodied Agents for Open-World Mobile Manipulation: REAL trains Qwen3-VL-8B to explore, ask when necessary, and then manipulate through a tool interface that withholds privileged object state; SFT followed by online GSPO reaches 56.9% success on the interactive REAL-Bench split, and the high-level policy transfers zero-shot to a physical robot with 78.3% success over 60 episodes.
From Gaze to Meaning: An AI Agent for Unified Zero-Shot Grounding and Explanation: Gaze Target Agent (GTA) renders predicted gaze as an arrow for a VLM, uses first-token confidence to trigger retrieval of labeled examples, and selects an object instance using the gaze heatmap, raising GazeFollow Acc@1 from GGVL's 0.450 to 0.493 without parameter updates, although its memory still uses training-set labels.
GEAR-Seg: A Grounded Explainable Agent for Reasoning Segmentation and Data Engine: GEAR-Seg converts an image into detailed descriptions of indexed regions, then lets a text-only LLM select masks according to the user's intent, achieving 57.5 ReasonSeg gIoU without task-specific fine-tuning and extending the same pipeline into an explanation-producing data engine for end-to-end student training.
GUIDE: Resolving Domain Bias in GUI Agents through Real-Time Web Video Retrieval and Plug-and-Play Annotation: GUIDE is a training-free, plug-and-play framework that automatically retrieves and extracts domain-specific planning and grounding knowledge from YouTube tutorial videos. By injecting this knowledge into corresponding modules of GUI agents to eliminate domain bias, it brings performance improvements of +4.47 to +7.48 percentage points for three agents with different architectures on OSWorld.
Guiding the Blind: Generalizing GUI Agents to Unseen Websites via Multimodal Tutorials: On WebOne's website-disjoint training and test split, WebLearner uses hierarchical memory and step-level GRPO rewards to learn to understand the current page, identify a relevant tutorial subtask, and execute an action, raising unseen-website task success from the Qwen3-VL-8B RAG baseline's 36.8% to 56.9%, a gain of 20.1 percentage points.
KATANA: Knowledge-Aligned Topology-Aware Neural Agents for RL-Driven Vision-Language Model Compression: KATANA treats the pruning algorithm itself as a learnable program, trains a code agent with rewards measured in a sandbox, and discovers KIRI's combination of cubic sparsity scheduling, Dual-Norm Activation scoring, and in-mask reconstruction; at 70% sparsity, LLaVA-1.5-7B reaches 84.6 MSCOCO CIDEr versus GSOP's 83.3, a 1.3-point gain, with approximately 2.8x decode acceleration on the specified sparse runtime.

Browse all 23 LLM Agent papers →

👥 Multi-Agent (3)¶

A Benchmark and Multi-Agent System for Instruction-driven Cinematic Video Compilation: CineBench evaluates instruction-driven cinematic compilation, while the training-free CineAgents system analyzes source footage before planning and executing an edit, achieving a shot-level F1 of 64.13, temporal correctness of 52.09%, and adversarial rejection of 87.23% on this benchmark.
AGE: Agentic Gaussian Editing in 3D Scenarios: AGE coordinates existing 3D editing tools through planning, execution, reflection, and backtracking agents with shared spatial memory, achieving a CLIP text–image direction similarity of 0.192 versus DGE's 0.133 on a custom benchmark, although component ablations are needed to isolate the sources of improvement.
Automatic Method Illustration Generation for AI Scientific Papers via Drawing Middleware Creation, Evolution, and Orchestration: FigAgent turns recurring visual components into callable, evolvable drawing functions and orchestrates them through multi-agent lookahead search to produce editable SVGs, improving node-alignment F1 on FigAgentBench from PaperBanana's 46.3 to 51.4.

🔒 LLM Safety (7)¶

Attention Misses Visual Risk: Risk-Adaptive Steering for Multimodal Safety Alignment: MoRAS improves multimodal risk assessment with a short image description, then uses the estimated risk to steer early response activations for the original query toward refusals, reducing LLaVA-1.5-7B's MM-Safety attack success rate from 40.1% to 2.6% while retaining its reported general-task aggregate scores.
BBQ-V: Benchmarking Visual Stereotype Bias in Large Multimodal Models: BBQ-V evaluates unsupported group-based inferences using real multi-person images and under-informative open-ended questions: its 4,497 images support 14,144 image–question pairs, and an input ablation shows that seeing more does not necessarily produce more cautious judgments.
Learning to Compose: Revisiting Proxy Task Design for Zero-Shot Composed Image Retrieval: FoCo (Focus-then-Complete) models vision-language composition in Zero-Shot Composed Image Retrieval (ZS-CIR) as a learnable two-stage process of "focusing then completing." By jointly training two proxy tasks—text-anchored visual aggregation and context-conditioned semantic completion—along with a cross-instance contrastive loss to prevent shortcut learning, it comprehensively outperforms existing methods across four ZS-CIR benchmarks without relying on LLMs during inference.
Neural Gate: Mitigating Privacy Risks in LVLMs via Neuron-Level Gradient Gating: This paper proposes Neural Gate, a neuron-level gradient gating method for mitigating privacy risks in LVLMs. The core mechanism is to first measure the consistency of each neuron's contribution to privacy targets on privacy subject features using a learnable vector, classifying neurons into three categories: strongly active, weakly active, and inactive. Then, during model editing, parameter updates are applied only to the gradients of strongly active privacy neurons, while gradients of other neurons are truncated. On MiniGPT and LLaVA, Neural Gate achieves a rejection rate of over 94% on sensitive queries with almost no loss in general task performance, while displaying robust generalization on out-of-distribution privacy categories.
ReShift: Aha-Moment-Driven Reasoning-Level Backdoor Attacks on Vision-Language Models: ReShift proposes a reasoning-level backdoor attack method targeting Vision-Language Models (VLMs). By leveraging the reinforcement learning-induced cognitive behavior of the "aha moment," it guides a controllable shift in the reasoning trajectory. This redirects the predicted results to a preset target answer while maintaining logical consistency, making it significantly harder to detect than traditional output-level backdoor attacks.
SlowBA: An efficiency backdoor attack towards VLM-based GUI agents: SlowBA proposes the first efficiency backdoor attack against VLM-based GUI agents. Through a two-stage Reward-level Backdoor Injection (RBI), the agent is trained to generate ultra-long responses during SFT, and a trigger-aware reward function is then used in RL to distinguish between triggered and clean inputs. Consequently, the agent exhibits extremely high response latency when encountering pop-up triggers, while the task accuracy remains largely unaffected.
Towards Benign Memory Forgetting for Selective Multimodal Large Language Model Unlearning: This paper proposes a new paradigm of "benign memory forgetting" and the S-MLLMUn Bench evaluation benchmark, and designs the SMFA (Sculpted Memory Forgetting Adapter) framework. By utilizing parameter masks guided by retaining anchors, it precisely erases privacy-sensitive knowledge in MLLMs without compromising the models' foundational visual understanding capabilities.

👻 Hallucination Detection (3)¶

Anatomy of a Lie: A Multi-Stage Diagnostic Framework for Tracing Hallucinations in Vision-Language Models: Cognitive Anomaly Detection (CAD) summarizes a vision-language model's evidence-to-answer process with three interpretable information measures and detects departures from a Gaussian-mixture model of nominal behavior, reaching a mean detection AUC of 0.858 on adversarial POPE while distinguishing perceptual instability, evidence-dependence anomalies, and decision uncertainty.
From Hallucination to Grounding: Diagnosing Visual Spatial Intelligence via CRISP: CRISP is a structural diagnostic benchmark for visual spatial intelligence. Through a dual-task paradigm of Spatial QA + 3D Scene Graph combined with a cross-task consistency protocol, it reveals whether VLM spatial reasoning represents genuine 3D geometric understanding or merely semantic shortcuts relying on language priors.
No Place to Hide: Benchmarking Video Hallucination with Background-Controlled Pairs: This paper proposes VidPair-Halluc, a video hallucination benchmark that utilizes adversarial video pairs with "highly similar backgrounds but significantly different foreground semantics" to decouple model errors from background variations and cleanly attribute them to foreground hallucinations. The accompanying PairFlow three-stage generation pipeline leverages T2I and video generation models to automatically synthesize 1K high-quality video pairs and 11K spatiotemporal QA pairs. Evaluations indicate that mainstream large video models generally suffer from a severe degradation in performance under this controlled setting.

📚 Pretraining (1)¶

GryphOne: Symbol-Aware Masked Diffusion for Structural Refinement in Offline Handwritten Mathematical Expression Recognition: GryphOne redefines handwritten mathematical expression recognition (HMER) from autoregressive sequence generation to an iterative symbol refinement process via discrete masked diffusion. It maintains syntactic consistency under local editing using Symbol-Aware Tokenization (SAT) and enhances refinement stability through Random-Mask Mutual Learning (RMML). This approach comprehensively outperforms re-implemented autoregressive baselines and commercial HMER systems on MathWriting, achieving a 5.51% CER and a 59.9% ExpRate.

💬 LLM (Other) (1)¶

SAM2Matting: Generalized Image and Video Matting: SAM2Matting proposes a decoupled "tracker-to-matting" framework that divides video matting into high-level tracking (where a frozen VOS tracker like SAM2/SAM3 ensures temporal consistency) and low-level matting (where a trainable ROI detector + progressive Alpha predictor estimates fine-grained transparency). Trained solely on image matting data, it achieves SOTA in a zero-shot manner on video matting benchmarks. It also supports multiple prompt types, maintains strong temporal consistency, and generalizes well to both human and open-world scenarios.

🔍 Information Retrieval & RAG (3)¶

CaPCL: Caption-Preserved Continual Learning for Text-to-Image Retrieval: When learning a new image-text retrieval task, CaPCL requires the model to reproduce auxiliary captions generated for current images before the update, protecting prior semantics omitted by new queries; averaged over four BLIP-2 task orders, it reduces ModX's forgetting measure FM from 0.0890 to 0.0654 while raising AA from 0.6387 to 0.6942.
HyFL-CLIP: Hyperbolic Fine-Tuning of CLIP for Robust Long-Context Understanding: HyFL-CLIP distills the Euclidean image-text alignment of pretrained CLIP into hyperbolic space (Lorentz model). By modeling the "global description - local component" hierarchical entailment relationship via Einstein midpoint aggregation, the model stably retrieves correct images even when long texts are perturbed (reordered, deleted, word-dropped), achieving up to a 19.5% improvement over the strongest baseline under word perturbations.
LightSTAR: Efficient Visual Document Retrieval via Lightweight Selection with Vision-Adaptive Refinement: To address the issue where MLLMs are too slow to encode all pages in visual document retrieval, LightSTAR proposes first filtering out candidates with high recall using an LLM-free lightweight visual selection module, and then performing vision-adaptive semantic refinement only on those candidates. This achieves 89.1 NDCG@5 on ViDoRe while reducing the end-to-end latency for 5,000 pages to 123.9s.

🎨 Image Generation (64)¶

A Scalable Vector Graphics Latent Space: SLS compresses each SVG path, including its geometry and style, into a decodable dense vector, preserving strong reconstruction quality while reducing the average captioning input from 2432.02 XML tokens to 15.56 path embeddings; its main contribution is a vector-graphics representation interface rather than a new text-to-SVG generator.
AccelAes: Accelerating Diffusion Transformers for Training-Free Aesthetic-Enhanced Image Generation: AccelAes uses aesthetic prompt semantics and cross-attention to identify regions deserving precise updates, combines sparse computation with spatial CFG and step-level prediction reuse, and reduces Lumina-Next latency from 12.37 to 5.86 seconds while increasing ImageReward from 0.7518 to 0.8410 without training.
Accelerated Likelihood Maximization for Diffusion-based Versatile Content Generation: ALM proposes a completely training-free sampling strategy for diffusion models. It performs explicit likelihood maximization optimization on unobserved regions during the reverse diffusion process, and leverages the property that "updates of adjacent iterations are approximately equal" to collapse \(N\) iterations into a single update. It significantly outperforms prior state-of-the-art methods at zero training cost across multi-modal tasks including image inpainting/outpainting, human motion completion, 3D texture generation, and long video generation.
Accelerating Diffusion Transformers with Gaussian Process Rectified Feature Cache: GP-Refiner treats full-compute features on an accelerated trajectory as noisy observations, corrects cache predictions with an online Gaussian process, and triggers recomputation through posterior uncertainty; on Qwen-Image, it raises PSNR from 28.58 to 29.48 while reducing computation relative to a fixed-interval TaylorSeer configuration.
Achieving Subcategorical Erasure in Text-to-Image Models: SURE starts from a single parent-category target, maps that category and retrieved related concepts toward an empty prompt, preserves neutral predictions, and regularizes local latent variation, reaching a 10.00% subcategory detection rate on Guns without establishing complete erasure of every subcategory.
ActionParty: Multi-Subject Action Binding in Generative Video Games: ActionParty jointly predicts a shared game video and each character's coordinates, separating action ownership through attention masks and grounding subjects through spatial RoPE; across 46 Melting Pot games with up to 7 players, it raises movement accuracy from Text-Action's 0.158 to 0.779.
AdaBridge-SR: Adaptive Bridge Matching for Real-World Image Super-Resolution: AdaBridge-SR predicts where to inject noise and when to enter high-variance bridge states from the degraded input, using shared weights for deterministic one-step super-resolution and stochastic few-step refinement: on RealSR, it reaches 25.2152 dB PSNR in one step and 0.6847 MANIQA in two steps, while additional steps still incur a fidelity cost.
Adaptive Noise Covariance Scheduling under Riemannian Metrics for Diffusion Models: The paper separates blue-to-white noise evolution into a covariance path and a per-image transition speed, combining a Bures-Wasserstein geodesic with a latent-predicted temperature to improve image-generation FID and KID without replacing the BNDM denoising backbone.
Advancing WordArt-Oriented Scene Text Recognition: Datasets and Methods: This paper systematically advances WordArt-oriented scene text recognition (WATER) from both data and model perspectives. On the data side, a 2M synthetic WordArt dataset named WATER-S is constructed, consisting of a rendering-based subset WATER-T and a generative-model-based subset WATER-Z. On the model side, WATERec is proposed, supporting arbitrary-shape input and autoregressive decoding. It achieves over 90% accuracy (90.40%) on WordArt-Bench for the first time, significantly outperforming general VLMs and OCR-specific VLMs.
Adversarial Score Distillation for Stable One-Step Diffusion in Real-World Image Super-Resolution: ASDSR reorganizes the generator's adversarial score guidance, computes the discriminator penalty on ground-truth latents, and adds lightweight CLIP supervision for one-step 4x real-world super-resolution, reaching 28.77 PSNR, 0.7973 SSIM, and 0.2896 LPIPS on DrealSR without leading every perceptual metric.

Browse all 64 Image Generation papers →

🎬 Video Generation (18)¶

Accelerating Text-to-Video Generation with Calibrated Sparse Attention: CalibAtt compiles input-stable block sparsity and spatial row repetition into two attention execution paths without updating model weights, reducing Wan 2.1 14B's 720p generation latency from 1244 to 785 seconds while retaining comparable VBench performance.
Aligning Human Sense: Calibrated Distributional Reward Learning for Video Generation: The paper connects consensus-based annotation calibration, multidimensional reward learning, and Wasserstein-constrained GRPO, improving action pairwise accuracy excluding ties from 80.30% to 83.33% and the reported aggregate scores of three video generators, although not every individual metric improves.
Anchor Forcing: Anchor Memory and Tri-Region RoPE for Interactive Streaming Video Diffusion: Anchor Forcing addresses evidence loss at prompt switches with anchor-guided re-caching and positional distribution shift during long rollouts with tri-region RoPE, improving the LongLive baseline from 26.37 to 73.00 in dynamic degree and from 78.92 to 82.25 in quality on interactive 60-second videos.
AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories: Instead of fusing historical views into one global point cloud, AnchorWeave retrieves complementary local point clouds, renders multiple anchor videos, and learns to reconcile them, improving its CogVideoX variant from 19.01 to 20.96 dB PSNR and from 0.6145 to 0.6727 SSIM over single-anchor conditioning under partial-revisit evaluation.
AnyFlow: Any-Step Video Diffusion Model with On-Policy Flow Map Distillation: AnyFlow replaces endpoint-only video distillation with transitions across arbitrary time intervals, then corrects student rollouts using three-segment shortcut simulation and distribution matching, reaching 84.05 VBench at 4 NFEs and 84.41 at 32 NFEs with its 14B causal model.
ART-VSR: Adaptive Rectified Trajectories for One-Step Video Super-Resolution: ART-VSR estimates how far each token should travel along a generative path, then rectifies its latent starting state accordingly, enabling one-step restoration with a Wan2.1 video prior; it achieves 0.197 LPIPS on REDS under the authors' unified re-evaluation protocol, without dominating every fidelity or perceptual metric.
Benchmarking Scientific Understanding and Reasoning for Video Generation using VideoScience-Bench: VideoScience-Bench places video generators in experimental scenarios requiring combinations of scientific concepts, evaluates their outputs with expert ratings and an evidence-grounded VideoScience-Judge, and shows that visual coherence is not scientific correctness, while checklist-based judging reaches a Spearman correlation of 0.96 with expert rankings.
Better Call CineCrew: Consistent Ultra-Long Narrative-to-Film Generation: CineCrew uses FilmDSL to organize narrative intent, cinematic staging, assets, and repair feedback into an executable production workflow without retraining video models, improving character, prop, and story continuity on 20 MovieBench narratives without leading every visual-quality or physics metric.
Beyond Inpainting: Unleash 3D Understanding for Stable Camera-Controlled Video Re-rendering: DepthDirector controls novel viewpoints with warped depth and preserves appearance and motion through the source video, avoiding erroneous textures inherited from warped RGB; its full model achieves RE 1.010, TE 0.053, and CamMC 1.434 under the absolute-camera protocol.
ConfCtrl: Enabling Precise Camera Control in Video Diffusion via Confidence-Aware Interpolation: ConfCtrl uses endpoint-conditioned video interpolation as a generative prior, changes the starting distribution of rectified flow with confidence-weighted projected point-cloud latents, and refines camera conditioning through predict-update modules, achieving 15.54 dB PSNR, 0.143 m translation error, and 0.103 rad rotation error on CO3D-Hydrant.

Browse all 18 Video Generation papers →

🧩 Multimodal VLM (21)¶

360° Image Perception with MLLMs: A Comprehensive Benchmark and a Training-Free Method: 360Bench exposes failures in fine-grained and spatial understanding of high-resolution panoramas, and training-free Free360 combines question-relevant entities, local attributes, inter-entity relations, and viewer directions into a scene graph, raising Qwen2.5-VL-7B overall accuracy from 38.1% with CMP input to 45.3%.
3D-Aware VLMs with Implicit and Explicit Geometries: VLM-IE3D injects both implicit geometric features and explicit tokens derived from reconstructed depth into Qwen2.5-VL, raising 3D video detection [email protected] from the backbone's 30.9 to 42.8, although its average spatial reasoning gain over an existing geometry-enhanced model is modest.
3D FaceShell: Attribute Transfer in 3D Face Avatars as a VLM Defense Mechanism: 3D FaceShell adds a learnable Gaussian shell to a frozen 3D face and uses multi-view semantic alignment to redirect VLM attribute predictions, reaching 48.7% injection rate on VideoLLaMA3 with LPIPS 0.1499 and identity similarity 0.7629 on frontal views; its contribution is the trade-off between semantic intervention and visual fidelity, not uniformly maximal attack success.
Actor as Its Own Critic: Unifying Region Understanding and Localization via CycleGRPO: CycleGRPO trains one multimodal model to describe a region and then localize its own description, using reconstructed-mask overlap to improve both directions without caption ground truths in its main training loop; SAMTok-4B improves from 61.9 to 67.7 on DLC-Bench and from 57.5 to 67.6 overall mask gIoU on GroundingSuite.
Adapting MLLMs for Nuanced Video Retrieval: TARA turns a generative MLLM into a unified retrieval encoder by fine-tuning its shared language model on just 20,000 carefully constructed text triplets, raising Tarsier 2's CiA-SSv2 Chiral mAP from 77.7 to 88.9 while also improving negated queries and composed video retrieval.
AMCI: Unlock the Potential of Large Multimodal Models for Fine-grained Open-world Classification via Adaptive Memory Context Injection: AMCI turns previously observed test images and their attribute descriptions into retrievable memory, then uses a language model to refine neighboring context into recognition prompts, improving Qwen2-VL 7B's mean CFI across five fine-grained datasets from 36.6 to 63.0 without updating parameters, while still depending on stream quality and additional language-model inference.
An Inverse-Adversarial and Difficulty-Adaptive Robust Vision-Language Model: IADA-RVLM constrains adversarial training with inverse-adversarial examples moved toward the correct class, learns specialized knowledge through visual prompts and a text adapter, and adjusts text knowledge fusion using class semantic distances, achieving the best Nat/Rob harmonic mean on novel classes across five datasets, including 52.52% PGD-100 robust accuracy on Caltech101.
AnchorGUI: Asymmetric Memory for Dual-Scale Learning in GUI Navigation: AnchorGUI uses cognitive state anchors to compare expected action effects with actual interface changes, compressing ordinary steps into text while retaining visual evidence for mismatches, so that one shared VLM supports both intra-trial correction and cross-trial distillation, achieving 57.3% single-attempt and 69.2% three-attempt cumulative success on AndroidWorld.
Art Beyond Semantics: Sheaf-Informed Contrastive Learning for Multi-Relational Representations: CANVAS modulates each artwork or text into relation-specific representations and uses a training graph's line-graph heat kernel to soften false-negative penalties, achieving the strongest image-to-text retrieval results across three art datasets without requiring full graph connectivity at inference.
ATOMIC: A Domain-Specific Vision-Language Model for Transmission Electron Microscopy: ATOMIC retains the LLaVA-v1.5 architecture, converts published TEM subimages and captions into complementary visual and domain supervision, and performs two-stage adaptation, reaching 75.2% TEM-MCQ accuracy and 37.5% TEM-VQA Answer Word Coverage with its 7B model, without yet replacing quantitative microscopy analysis.

Browse all 21 Multimodal VLM papers →

🧠 VLM Reasoning (19)¶

3D-Layout-R1: Structured Reasoning for Language-Instructed Spatial Editing: 3D-Layout-R1 trains models to update JSON 3D scene graphs while reasoning, then refines their outputs with IoU, collision, and format rewards through GRPO, raising Qwen3-8B's mean IoU on perfect-input sorting from 0.850 with Vanilla GRPO to 0.924, although gains and metric trade-offs depend substantially on the task.
A4-Agent: An Agentic Framework for Zero-Shot Affordance Reasoning: A4-Agent assigns interaction imagination, functional-part selection, and pixel-level localization to specialized pretrained models, reaching 70.52 gIoU on ReasonAff without affordance-specific fine-tuning, at approximately 32 seconds per inference for the full pipeline.
AdaThinking-E: One-Token Entropy Regulation for Adaptive Thinking: AdaThinking-E optimizes the first token that decides whether to think separately from the subsequent response, using high-entropy exploration followed by low-entropy convergence to reach 26.5 on OCR-Reasoning with its Qwen2.5-based model while matching approximately R-4B's CharXivRQ performance with only 72 output tokens on average.
AnE: Pushing the Reasoning Frontier of Multimodal LLMs via Anchor Evolution: AnE retrieves real problems around a student's persistent failures, uses teacher hints to elicit the student's own successful reasoning trajectories, and removes external assistance through hint-free reinforcement learning, raising Qwen2.5-VL-7B's average accuracy across eight benchmarks from 51.6% to 61.9% after three evolution rounds.
Attention-based Vision-Language Memory for Spatial Reasoning: VLM² aligns semantic features, predicted geometry, and viewpoint information from monocular video before using sliding-window working memory and fixed-capacity episodic memory to retrieve, fuse, and update scene representations, reaching 68.8 on VSI-Bench with video-only inputs, 7.9 score points above VLM-3R.
Before Thinking, Learn to Decide: Proactive Routing for Efficient Visual Reasoning: PRP trains a small vision-language model to predict its own and a larger target model's solvability before lengthy reasoning begins; on MathVista, JRL keeps 60% of requests with the draft and achieves 74.2% accuracy at 26.9 seconds average latency, versus 74.1% and 64.8 seconds for target-only inference, with a reported 2.41x speedup.
Benchmarking Dynamic Affective Reasoning: A Viewer-Centric Video Emotion Dataset: DAR redefines video emotion understanding through event boundaries, 27 viewer-emotion categories, and contextual causal explanations, while DAR-R1, obtained by applying SFT and GRPO to Qwen2.5-VL-3B, achieves 41.5% segment-count accuracy, 52.3% temporal mIoU, and 28.6% emotion accuracy on its test set.
Beyond Final Answers: CRYSTAL Benchmark for Transparent Multimodal Reasoning Evaluation: CRYSTAL adds checkable reference reasoning chains to 6,372 visual questions, exposes incomplete reasoning behind correct answers through content coverage and relative-order evaluation, and raises Qwen2.5-VL-3B's Match F1 from 0.480 to 0.633 and accuracy from 39.85% to 47.52% with CPR-Curriculum.
Beyond Where to Look: Trajectory-Guided Reinforcement Learning for Multimodal RLVR: TGRL mixes a small number of complete expert reasoning trajectories into the student's RLVR rollout groups, using off-policy probability-ratio correction, adaptive token reweighting, and successful-trajectory filtering to connect observed facts to reasoning, improving Qwen2.5-VL-7B's five-benchmark DAPO average from 58.15 to 60.68 in the 39K-data setting.
BrepCoder: A Unified Multimodal Large Language Model for Multi-task B-rep Reasoning: BrepCoder aligns B-rep surface geometry and topology with Python-like CAD programs, learns shape-to-construction correspondences through reverse engineering, and adapts to completion, correction, and QA, reducing median Chamfer distance from CADCL's 0.972 to 0.464 on DeepCAD reverse engineering, both on the \(10^{-3}\) scale, without leading every task or metric.

Browse all 19 VLM Reasoning papers →

⚡ VLM Efficiency (18)¶

3DZip: Spatial-Aware Feature Diversity-Guided Token Compression for 3D Question Answering: 3DZip compresses LLaVA-3D visual tokens before the language model through coarse voxelization, feature-diversity DPP anchor selection, and spatially constrained merging, reducing the reported 1410-token configuration to 128 tokens while retaining a paper-reported 94.7% relative performance and reducing latency from 342 ms to 178 ms per sample on a single RTX 4090.
Accelerating Diffusion Models via Equal-Risk Caching: ERC probes how a single cache-window length affects final outputs at different timesteps, accumulates the resulting one-dimensional risk curve, and partitions it to allocate refreshes at lower offline cost than exhaustive interval evaluation, achieving, for example, 3.23x acceleration and LPIPS 0.1954 on FLUX.1 [dev] with a refresh budget of 15.
Activation Quantization of Vision Encoders Needs Prefixing Registers: RegCache extracts reusable middle-layer register KV caches from reference images and combines prefix insertion with residual outlier token deletion to improve post-training quantization, raising CLIP-B/16 ImageNet-1k accuracy under ERQ W4A4 from 1.56% to 46.07%, without universally recovering full-precision performance.
AnchorPrune: Relevance-Anchored Contextual Expansion for Visual Token Pruning: AnchorPrune separates visual token selection into the ordered stages of protecting query-relevant evidence and adding important, non-redundant context, requiring no training and retaining 97.6% of full-model average relative performance on LLaVA-NeXT-7B with only 160/2880 tokens.
ARVAR: Accelerating Visual Autoregressive Model via Attention Retrospect: ARVAR uses low-resolution attention and activations to allocate high-resolution computation, recomputing important tokens, reconstructing other positions, and reusing the conditional branch at selected late steps to achieve 1.56x HART and 3.01x Infinity-2B speedups on one RTX3090 while preserving the original model's images better than FastVAR.
AViTS: Adaptive Spatiotemporal Token Selection for Efficient Dynamic-Resolution Generation: AViTS jointly asks which latent tokens are relevant to the text and which are still changing during denoising to decide where to upsample early, reducing FLUX.1-dev latency from 25.78 s to 3.65 s in its 14-step configuration while ImageReward changes from 0.9719 to 0.9723, although more aggressive acceleration still sacrifices quality.
BATQuant: Outlier-Resilient MXFP4 Quantization via Learnable Block-wise Optimization: BATQuant confines learnable non-orthogonal transformations to MXFP4's 32-element quantization blocks, then uses shared factorization and local clipping to reduce deployment overhead and residual outlier effects, achieving 96.43% multimodal performance recovery on Qwen3-VL-8B-Instruct under W4A4KV16.
Bayesian Self-Attention with Local Pixel Correlations for Lightweight Denoising Transformers: NLformer extends lightweight denoising Transformers from finding and averaging similar patches to also exploiting within-patch dependencies through explicit inverse-based processing, reaching 39.88 / 39.92 dB on SIDD / DND with 2.44M parameters and outperforming the lightweight models compared in the paper, without minimizing every resource metric.
BLOB-Q: Boosting Low Bit ViT Quantization via Global Optimization on Model Distortion: BLOB-Q turns mixed-precision allocation for ViT weights and activations into an additive approximation of model-output distortion, then uses dynamic programming to find the best combination within a bit budget, reaching 76.85% ImageNet top-1 on ViT-S at 4-bit average precision; its global-optimality guarantee should be restricted to the approximated discrete problem, not unconditional optimal accuracy for the original network.
CASA: Cross-Attention over Self-Attention for Efficient Vision-Language Fusion: CASA revisits cross-attention through controlled experiments, keeping image representations outside the language backbone and replacing window-local visual KV sources for efficient fusion; vanilla CA reaches 6.8 FPS versus 1.2 FPS for token insertion in the controlled Helium1 streaming setup, but complex infographic understanding still shows a substantial performance gap.

Browse all 18 VLM Efficiency papers →

🎵 Audio & Speech (5)¶

LipsFlow: First Exploration of Neuromorphic OT-CFM in Multi-Speaker Visual Speech Recognition: LipsFlow for the first time introduces neuromorphic event camera perception and optimal transport conditional flow matching (OT-CFM) to multi-speaker visual speech recognition. By capturing millisecond-level lip dynamics through learnable event representations, implementing deterministic 2-step ODE decoding with OT-CFM, and resolving homophenous ambiguities via dual-level semantic supervision, it achieves state-of-the-art performance on DVS-Lip with a 22.3% WER and 240ms latency.
MG-RWKV: Multi-Grained Context-Aware RWKV for Temporal Forgery Localization: MG-RWKV proposes a temporal forgery localization (TFL) framework based on the linear-complexity recurrent architecture of RWKV. By capturing global temporal context via bidirectional RWKV, adaptively selecting explicit temporal receptive fields through a Multi-Grained Mixture of Experts (MG-MoE), and eliminating contradictory predictions across multi-scale features using a Cross-Grained Consistency constraint (CGC), it comprehensively outperforms previous state-of-the-art methods on four benchmarks: Lav-DF, TVIL, Psynd, and AV-Deepfake1M, while maintaining \(O(T)\) linear complexity with an inference time of only 73.4ms.
See & Sniff: Learning Visuo-Olfactory Representations: Using the key insight that "odor identity remains invariant to visual transformations within semantic categories," this paper extends the smell-only SmellNet dataset into a paired visuo-olfactory dataset, SmellNet-V. It designs a self-supervised dual-stream framework named See & Sniff based on dense local contrastive alignment. The proposed method significantly outperforms smell-only baselines in odor classification, cross-modal retrieval, and a newly introduced pixel-level odor localization task.
Sparsity-Inducing Divergence Losses for Biometric Verification: Q-Margin shifts the margin penalty from geometric logit modifications to a probabilistic reference measure in \(\alpha\)-divergence losses. While maintaining posterior sparsity, it consistently outperforms ArcFace/CosFace baselines in low-FAR scenarios on IJB-B/C and VoxCeleb, and reduces training throughput overhead from 27% to 5% via exact top-K truncation.
Step-by-Step Video-to-Audio Synthesis via Negative Audio Guidance: Addressing the limitation where existing video-to-audio (V2A) models can only generate a complete audio track all at once instead of adding sounds layer-by-layer like Foley artists, this paper proposes Negative Audio Guidance (NAG). By training a branch conditioned on "already-generated audio tracks" and utilizing it inversely during sampling, the current generation is pushed away from existing sounds. This enables step-by-step generation of complementary audio tracks to mix into high-quality synthesized sound, using only standard single-reference audio-visual datasets.

🔎 AIGC Detection (3)¶

Asymmetric Anchoring: Opening the Black Box of MLLMs for Forgery Detection: AAP aligns intermediate multimodal representations with a frozen visual reference only for real images, then uses patchwise alignment errors to guide localization, achieving 97.1% detection accuracy and 56.8% localization IoU with AAP-lisa on SID-Set.
Efficient Document Tampering Localization with Multi-Level Discrepancy Features and Unified DCT-Quantization Embedding: DiffNet introduces two complementary designs: a multi-level discrepancy transform that converts the feature pyramid from content signals into sign-invariant discrepancy intensity signals, and a unified DCT-quantization joint embedding that replaces the traditional, high-overhead Frequency Perception Head (FPH) with discrete embeddings. This achieves approximately a 30% F1 improvement on cross-domain and human-manipulated document localization at a lower computational cost, with up to a 7-fold increase in throughput.
Trustworthy Image Authentication using Forensic Knowledge Graphs: This work combines the "evidence-seeking" capability of image forensics with the natural language generation of VLMs into a single system: first, a self-supervised forensic fingerprint backbone is used to segment the image into forensically consistent regions and predict the source identity, post-processing history, and compression parameters for each region to construct a structured "Forensic Knowledge Graph (FKG)"; subsequently, a VLM is guided to generate step-by-step verifiable forensic explanations based strictly on the graph, consistently outperforming pure forensic models and pure VLMs in terms of detection, forgery typing/localization, and forensic explanation.

🧊 3D Vision (85)¶

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-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.
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.
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 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.
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.
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.

Browse all 85 3D Vision papers →

🎯 Object Detection (2)¶

A Simple Baseline with Placement Prior for Point-Supervised Oriented Object Detection: PlacePriorBase combines neighborhood angle alignment, image-level size regularity, and training-free watershed box targets to build an inexpensive point-supervised detector, while its transferable losses improve stronger detectors without making the lightweight baseline itself the most accurate model.
Back-Tracking from Clarity: Self-Learning to See Text from Afar: The method propagates reliable detections of clear text backward into earlier, degraded video frames, combining point tracking with carrier geometry to train an unchanged single-frame student; its SceneText50 tracking AUC reaches 44.71 versus 38.65 for CoTracker3.

✂️ Segmentation (2)¶

Beyond Aesthetics: Quantifying Information Loss in Turbid Scenes: The paper introduces the controlled TUB dataset and Phase-Congruency Delentropy (PCD) to distinguish unattractive underwater images from images that have lost task-relevant structure; PCD correlates most strongly with segmentation performance in the reported real and synthetic experiments, but is not a universal measure of recognizability across arbitrary scenes.
Bridging the Geometry Mismatch: Frequency-Aware Anisotropic Serialization for Thin-Structure SSMs: FGOS-Net separates low-frequency topology modeling from high-frequency detail filtering, using deterministic sub-band-matched scanning and topology-conditioned active probes to reduce crack fragmentation and texture false positives; it achieves 91.29% mIoU and 97.12% clDice on DeepCrack, with 80.2 FPS under the paper's specified RTX 3090 evaluation setting.

🖼️ Image Restoration (13)¶

340 FPS Reflection-free Video from Spikes Modulated by a Rapidly Rotating Polarizer: A rapidly rotating polarizer modulates spike-camera measurements, and reliability-weighted physical separation followed by residual refinement reconstructs reflection-free video at a nominal 340 FPS output cadence, achieving 27.0598 dB PSNR on the authors' synthetic test set.
CogSENet: Blind Image Deblurring with Blur-Conditioned Semantic Routing and Explicit Frequency Fusion: CogSENet reformulates blind image deblurring from a passive pixel regression into an active, semantically aligned reconstruction process. By mimicking the active visual scanning, retinal functional differentiation, and focal adaptation of the eagle visual system, it proposes three core designs: a Semantic-Driven State Space Module (SDSSM), a Bi-Frequency Fusion Block (BFFB), and joint modulation based on a Continuous Blur Field (CBF) and CLIP semantics. With only 8.9M parameters, it outperforms state-of-the-art (SOTA) methods such as EVSSM and FFTformer on the GoPro, HIDE, and RealBlur datasets.
Fabric Image Demoiréing Benchmark from Synthesis to Restoration: This paper presents the first systematic study on the fabric image demoiréing problem. It introduces PRISM, a physical imaging chain simulation-based residual injection synthesis framework (including the first fabric moiré benchmark with 16,050 image pairs), and FaDeNet, a conservative restoration network specifically designed for the spectrally entangled characteristics of fabrics, significantly outperforming existing screen demoiréing methods in terms of PSNR, SSIM, and LPIPS.
Fidelity- and Perception-Aware Local Implicit Attention for Arbitrary-Scale Image Super-Resolution: FPLIA proposes a dual-stream framework that fuses the fidelity features of a regression backbone and the perceptual features of a diffusion model via asymmetric bidirectional cross-attention (FPAM) and pixel-wise adaptive selection (FPSM), simultaneously achieving high fidelity and high perceptual quality in ASISR.
FlowDec: Temporal Conditional Flow Decorruptor for Robust Continuous Vision-Language Navigation: FlowDec proposes an image decorruption framework based on conditional flow matching (CFM). By incorporating a hybrid temporal conditioning strategy and action centroid-guided filtering, it enhances robustness against diverse visual corruptions without modifying the downstream VLN backbone. It improves the relative navigation Success Rate (SR) of continuous VLN agents by 25.33% on R2R-CE and 9.38% on RxR-CE, while operating 3 to 8 times faster during inference compared to diffusion-based TTA baselines.
FMA-Net++: Motion- and Exposure-Aware Joint Video Super-Resolution and Deblurring: FMA-Net++ proposes a non-recurrent sequence-level framework based on Hierarchical Bidirectional Aggregation (HRBA) blocks. By introducing an Exposure Time-aware Modulation (ETM) module to inject frame-level exposure information into features, and combining it with an exposure-aware Flow-Guided Dynamic Filtering (FGDF) to jointly model motion- and exposure-varying degradations, the method achieves SOTA accuracy and inference speed on the VSRDB task.
FreeMEF: A Flexible Multi-Exposure Fusion Transformer for Arbitrary Number of Frames: FreeMEF proposes a "There and Back Again" two-stage paradigm: it first recursively aggregates multi-exposure features from an arbitrary number of frames into a global representation using a Recurrent State Space Module, and then restores the reference frame guided by an Extreme-Aware Hybrid Attention. This enables flexible inference on 2/3/5 frames while significantly suppressing ghosting artifacts and enhancing dynamic range.
FreqOrtho-SR: Frequency-Guided Orthogonal Expert Learning for Real-World Image Super-Resolution: FreqOrtho-SR proposes two core modules: Frequency-Guided LoRA Mixture-of-Experts (FreqMoE) and Orthogonal Gradient Projection (OGP). Driven by adaptive expert routing based on FFT degradation features and orthogonal constraints on pixel-semantic subspaces, it achieves degradation-adaptive real-world image super-resolution within a single-step diffusion model framework, obtaining optimal or sub-optimal trade-offs between fidelity and perceptual quality across multiple benchmarks.
LogicIR: Logic Gate Networks for Image Restoration: LogicIR is the first logic gate network (LGN) specifically designed for image restoration. Through a fully logic-gate UNet architecture, a differentiable bit decoding layer, and an Index Shuffling cross-group communication mechanism, it achieves competitive restoration quality on denoising, deblocking, and deraining tasks with operations (BOPs) far lower than BNN and LUT methods, demonstrating the feasibility of pure logic gate operations in image restoration.
Personalization as Inverse Planning: Learning Latent Design Intents for Agentic Slide Generation via Structural Denoising: Spire models page-level presentation slide personalization (PSP) as an inverse planning problem. By applying discrete structural perturbations to golden slides, it constructs self-supervised "denoising" signals to train Critic and Planner agents (7B-level) to cooperatively and iteratively refine design specifications under RL. Without relying on specific executors, it infers the user's implicit design intent and significantly outperforms strong GPT-based baselines in both visual similarity and VLM-as-Judge scores.

Browse all 13 Image Restoration papers →

🛰️ Remote Sensing (3)¶

Align and Segment: Unsupervised Learning for Building Segmentation From Misaligned Labels: Align and Segment (AnS) jointly learns building segmentation and geometric label correction, using self-supervised consistency and paired geometric augmentation to prevent memorized offsets, raising segmentation IoU from 0.39 for naive joint training to 0.78 on systematically misaligned Las Vegas data without clean aligned training labels.
Beyond 2D Matching: A Unified Single-Stage Framework for Geometry-Aware Cross-View Object Geo-Localization: GAGeo adapts the 3D foundation model π3 using CMA-Loc building correspondences and camera annotations, jointly decodes prompted objects and cameras in one forward pass, and uses satellite-anchored contrastive learning to reach 43.86 mAcc and 57.88 mIoU on the untrained ground-to-drone configuration.
Beyond Attention: Convolutional Global Context for Remote Sensing Change Detection: ChangeGCC conditions convolution kernels on scene-level descriptors, then combines intra-frame cross-scale aggregation with symmetric cross-temporal gating to detect remote sensing changes without a global pairwise attention matrix; its Base variant achieves 85.28% IoU on LEVIR-CD with 13.60M parameters.

🔍 Anomaly Detection (5)¶

A Comprehensive Analysis about Unsupervised Outlier Detection for Images: VUOD obtains reliable inlier/outlier pseudo-labels through fine-grained clustering and distance-ranking agreement, learns a lightweight feature mapping, and applies OCSVM to the improved representation, raising aggregated CIFAR-10 AUROC from FlexUOD's 0.942 to 0.972 while extending the same framework to industrial and medical images.
Anomaly Factory 3D: A Modular Framework for Diverse Pseudo-Anomaly Synthesis in Unsupervised 3D Anomaly Detection: AF3AD is a standalone geometric pseudo-anomaly generator that combines local coordinates, spatial kernels, directional gating, and stochastic presets; with a modified offset-prediction detector, it reaches object-level AUROC of 91.5% on AnomalyShapeNet and 85.2% on Real3D-AD.
ArcAD: Anomaly-Rectified Calibration for Cold-Start Supervised Anomaly Detection: ArcAD calibrates the latent space of a reconstruction-based industrial anomaly detector using balanced normal prototypes, prototype-filtered synthetic anomalies, and a few real defects, raising Dinomaly's image-level AUROC from 88.8% to 92.5% on multi-class cold-start Real-IAD while retaining its original reconstruction-based inference score.
BAAF: Universal Transformation of One-Class Classifiers for Unsupervised Image Anomaly Detection: BAAF cleans contaminated training images through disjoint bags, out-of-bag anomaly scoring, and majority voting before retraining one unchanged one-class detector, raising average image AUROC across eight detectors from 0.762 to 0.933 on MVTec AD with 10% training contamination without adding test-time models.
Beyond Common Sense: Grounding Logical Anomaly Detection in Inspection Criteria: SCAN turns industrial logical anomaly detection into explicit rule-by-rule visual inspection, combining FLAW training data, structured supervised fine-tuning, and rule-level reinforcement learning to raise an 8B model's strict scoring accuracy on MVTec-LOCO from 56.73% to 64.58%, while leaving substantial missed-detection concerns.

🧑 Human Understanding (19)¶

ActionPlan: Future-Aware Streaming Motion Synthesis via Frame-Level Action Planning: ActionPlan generates temporally aligned semantic action plans before progressively synthesizing 3D human motion, combining offline generation and streaming output in one model and reducing streaming FID from MotionStreamer's 11.790 to 5.735 on the HumanML3D-272 test set.
ANFI: Rethinking Neighbor Feature Interaction in Person Re-ID: ANFI learns both affinity and discrepancy interactions between neighboring person images, trains their sample-wise fusion with noisy relation supervision, and achieves 88.1% mAP on CUHK03 in the same-backbone comparison, 1.2 percentage points above its strongest comparator.
BackTranslation2.0 -- A Linguistically Motivated Metric to Assess Sign Language Production: Proposes BackTranslation2.0, a linguistically motivated text-to-sign language translation evaluation metric. Utilizing a two-phase agent framework (10 dedicated tools for evidence extraction + 4 LLM cross-comparison modules), it produces deterministic scores across four dimensions: grammatical correctness, phonological accuracy, motion fluency, and generation fidelity. It achieves a Pearson \(r=1.00\) / Spearman \(\rho=1.00\) correlation with human judgment on known corruptions and synthetic data, significantly outperforming existing backtranslation and motion similarity baselines.
BiCE-HG: A Bi-Conditional Egocentric Hand Gesture Dataset for Intelligent Reality Systems: BiCE-HG crosses three mobility states with two lighting configurations and provides spatial illuminance records alongside hand skeleton data; its dynamic baseline reaches 89.24% validation accuracy, but the evaluation is not participant-independent and does not establish cross-lighting generalization.
Dress-ED: Instruction-Guided Editing for Virtual Try-On and Try-Off: Dress-ED constructs the first large-scale benchmark dataset (146k verified quadruplets) that unifies VTON (virtual try-on), VTOFF (virtual try-off), and text instruction-guided garment editing. It proposes a unified multimodal diffusion baseline model, Dress-EM, based on MLLM + dual-path connector + DiT, which thoroughly outperforms general-purpose editing models and domain-specific VTON models on instruction-guided garment editing tasks.
FaceMoE: Mixture of Experts for Low-Resolution Face Recognition: FaceMoE replaces the single FFN in Transformers with multiple sparsely activated MoE experts and a Top-k router, allowing different experts to automatically specialize in distinct semantic regions of the face (high-frequency texture, low-frequency smooth, landmarks). This achieves resolution-aware feature extraction. It comprehensively outperforms the state-of-the-art (SOTA) on three low-resolution benchmarks (BRIAR, IJB-S, TinyFace) while suffering almost no degradation in high-resolution (HR) pre-trained performance.
FlowerDance: MeanFlow for Efficient and Refined 3D Dance Generation: FlowerDance combines MeanFlow (a flow matching variant that replaces instantaneous velocity prediction with interval-average velocity) with Physical Consistency Constraints (PCC), a BiMamba backbone, and channel-level cross-modal fusion. It generates high-quality 3D dance motions in only 5–20 sampling steps. While achieving state-of-the-art (SOTA) quality on FineDance and AIST++, its inference speed (2008 FPS) far exceeds the previous best method (MatchDance at 345 FPS), leaving ample computational budget for real-time 3D rendering.
InterEdit: Navigating Text-Guided 3D Dyadic Human Motion Editing: InterEdit proposes a new task, Text-guided 3D Dyadic Motion Editing (TMME), builds the first large-scale dyadic motion editing dataset InterEdit3D (comprising 5,161 source-target-text triplets), and designs the InterEdit method based on conditional diffusion models. By utilizing semantic-aware plan token alignment to capture high-level editing intentions and interaction-aware frequency token alignment to constrain interaction rhythm using DCT frequency band energy, InterEdit significantly outperforms four baseline methods in editing faithfulness and motion realism.
Multi-scale Object-Aware Gaze Estimation via Geometric Reasoning: This work reformulates gaze target estimation from pixel-level regression to a hierarchical reasoning problem. It first establishes candidate attention objects using object-level semantic representations, constructs a field-of-view (FOV) cone geometric prior using gaze direction to constrain the search space, and finally achieves precise localization via multi-scale residual fusion. This method achieves state-of-the-art (SOTA) performance on GazeFollow, VideoAttentionTarget, ChildPlay, and GOO-Real using only 7.1M parameters.
Odoriko: A Shape-Aware Multimodal Diffusion Framework for Human Motion: Odoriko proposes the first unified, multimodal human motion generation framework. By hierarchically injecting gender and SMPL shape parameters as explicit conditioning signals into the diffusion backbone, the generated motion reflects the biological morphological characteristics of the subjects. Meanwhile, it achieves or surpasses the performance of current task-specific methods across three tasks (text-to-motion, music-to-dance, and video-to-motion estimation) with a minimal parameter footprint.

Browse all 19 Human Understanding papers →

📹 Video Understanding (20)¶

A Dual-Transformer Architecture with Cross-Attention for Multi-Camera View Recommendation: The paper separates the encoding of previously shown footage from the evaluation of current camera candidates, letting candidates query temporal memory through cross-attention; with frozen SwinV2-Tiny features and focal loss, it reaches 69.65% [email protected] on TVMCE, a thresholded recommendation metric rather than forced six-way selection accuracy.
Adaptive Latent Trajectory Anchoring for Action Segmentation Dataset Condensation: The method compresses action-feature sequences into sparse DDIM latent anchors and allocates anchors by reconstruction error, reaching 68.0% ASFormer accuracy on Breakfast versus 62.8% for GNI at the same temporal anchor budget, while remaining below the 70.7% full-data result.
Aligning Anything: Hierarchical Motion Estimation for Video Frame Interpolation: The paper injects SAM object regions into contextual feature extraction, flow distillation, and feature supervision, retaining pixel-level motion flexibility while improving object-level coherence; under matched retraining settings, AMT-G improves from 36.42 to 36.62 dB on Vimeo90K.
Amplify, Aggregate, and Adjust: VideoMAE-based Holistic-Subtle Aggregation for Micro-Action Recognition: A3-MAE extends an RGB-only VideoMAE by locating and amplifying subtle-motion regions, exchanging holistic and local information bidirectionally, and calibrating distillation supervision with instance-wise confidence, improving MA-52 Action Top-1 from 66.47% to 67.92% and iMiGUE from 66.44% to 67.08%.
Bayesian Uncertainty Attribution-Guided Fine-Tuning for Open-Set Action Recognition: The paper turns ensemble uncertainty from a test-time rejection score into a representation-refinement signal, using three-stage fine-tuning and two-phase distillation to reach 88.50 AUROC with a single TPN student on UCF101/HMDB51, compared with 88.62 for the refined ensemble.
Bounding-Box Trajectories Matter for Video Anomaly Detection: TrajVAD promotes bounding-box trajectories already produced by detection and tracking to the primary anomaly signal, learning normal motion with a class-aware normalizing flow; its trajectory-only variant reaches 87.7 AP on ShanghaiTech, while reliability-gated pose fusion reaches 88.6 AUROC and 90.9 AP, although pose does not help on every dataset.
Cambrian-P: Pose-Grounded Video Understanding: Cambrian-P adds per-frame camera pose regression to video MLLM training and reconciles geometry learning with question answering through interleaved training and frame sampling jitter, improving VSI-Bench from 69.2 to 73.7 against a matched no-pose baseline; the gains primarily reflect better learned representations rather than inference-time pose-token use.
ChronusOmni: Improving Time Awareness of Omni-Modal Large Language Models: ChronusOmni interleaves absolute-time text, video frames, and corresponding audio for Ola-7B, then learns six temporal tasks through audiovisual dense-captioning SFT and task-reward GRPO, reaching 79.85 audio-to-time [email protected] on ChronusAV and 34.5 zero-shot temporal-grounding mIoU on LongVALE.
ClearText-Video: A Large-Scale Text-Centric Video Dataset Bridging Video Restoration and Scene-Text Enhancement: ClearText-Video (CTVid) connects low-level restoration with text-grounded video QA through high-quality, degraded, and restored versions of the same content, using 4,639 videos to show that sharper-looking inputs need not be read more accurately and that in-dataset fine-tuning can improve accuracy while reducing an uncertainty-aware metric.
CLUE-VAD: Structured Semantic Clues for Understanding Explainable Events in Video Anomaly Detection: CLUE-VAD organizes video evidence into Action, Environment, and Object captions, performs weakly supervised detection through category-aware fusion, and generates explanations using segment attention and a clue-conditioned language model, achieving 89.47% AUC / 88.24% AP on UCF-Crime / XD-Violence without establishing superiority over every multimodal detector.

Browse all 20 Video Understanding papers →

🚗 Autonomous Driving (52)¶

2K Retrofit: Entropy-Guided Efficient Sparse Refinement for High-Resolution 3D Geometry Prediction: This paper proposes 2K Retrofit, a universal framework enabling existing 3D foundation models (e.g., Depth Anything, VGGT) to output 2K resolution geometry predictions without modifying or retraining the base models. The core idea is to perform fast coarse prediction with a frozen base model and apply full-resolution correction only to high-uncertainty pixel regions using an entropy-guided sparse refinement mechanism, achieving or exceeding the accuracy of full-resolution methods while significantly reducing computational and GPU memory overhead.
A Physics-Grounded Benchmark for Multi-Agent Dynamics in World Models: CrashTwin combines traffic-interaction data, calibration-free reconstruction, and separate physical diagnostics to show that visually coherent world-model videos can still contain incorrect momentum exchange and kinetic-energy increases; its contribution is an evaluation framework, not a new video generator.
Adaptive Neural Dynamics for Robust Geometric LiDAR-Inertial State Estimation on UAVs: NeuroSDE-LIO retains the geometric estimation backbone of FAST-LIO2, adaptively combines linear bias priors with neural stochastic bias dynamics, and propagates their uncertainty into the filter, reducing UAV-Lab absolute pose error from 2.13 m to 1.34 m at a LiDAR processing rate of 19.2 Hz.
Agent-OBJ: Prompt-Driven 3D Adversaries for Multi-Modal Perception: Agent-OBJ searches language prompts controlling the shape and appearance of a 3D object using only detection boxes and confidence scores, combining naturalness and cross-view stability constraints to achieve 64.0% joint disappearance from camera and fusion detectors in simulated nuScenes insertion experiments.
AiSCREAM: Absolute Target Localization with Language-Conditioned Cross-View Alignment for Autonomous Vehicles: AiSCREAM uses an aerial image generated from a single front view as an auxiliary spatial cue, aligns the two views through language-conditioned semantic segmentation, and combines multitask localization with MLLM refinement to predict a metric target and acceptable region, reducing DRAMATiST RMSE from the strongest model baseline's 8.56 m to 5.00 m.
Analytic Bayesian Uncertainty for LiDAR Segmentation: A Single-pass Generative Approach: GMM-NIG places conjugate Bayesian posteriors over the means and variances of a class-conditional Gaussian mixture head attached to a deterministic segmentation backbone, then uses analytic Student-t prediction to reach 65.71% mIoU, 2.01% ACE, and 53.33% failure-detection AUPR with RangeFormer on SemanticKITTI without repeated stochastic forward passes.
ARC-Loc: Leveraging Azimuthal Ray Convergence as a Geometric Cue for Direct Cross-View Localization: Under a known heading or a bounded-noise heading prior, ARC-Loc converts ground-satellite matches into azimuthal lines and estimates position through differentiable weighted intersection and RANSAC, requiring neither depth estimation nor BEV transformation and reaching 2.52 m mean error on VIGOR Same-Area with known orientation, while remaining limited by narrow-view cross-area generalization.
ASSCG: Just-Right Gating over Chattering for Fast–Slow LLM Planning in Autonomous Driving: ASSCG learns not only when to query a slow language model but also when to reuse or suppress its cached guidance, using temporal gating and compute-aware reinforcement learning to raise AsyncDriver's nuPlan Hard20 score from 65.00 to 67.28 while reducing average end-to-end inference latency from 0.80 to 0.32 seconds per frame.
ASTAD: Asymmetric Style Transfer for Synthetic-to-Real Domain Adaptation in Autonomous Driving: Addressing the asymmetric constraint in autonomous driving scenarios where "synthetic images possess perfect annotations, whereas real-world style reference images are unannotated," ASTModel proposes a two-stage training-free framework. It first extracts coarse semantic priors from unlabeled style images via DINO prototype matching. During the reverse diffusion process, these priors are refined through multi-layer semantic voting, followed by class-consistent style injection enabled by robust median-threshold filtering and pixel-proportion modulated hybrid AdaIN. This generates realistic target-domain style data while preserving semantic structures, achieving a 3.2x speedup.
AutoWeather4D: Autonomous Driving Video Weather Conversion via G-Buffer Dual-Pass Editing: AutoWeather4D extracts metric geometry and materials from driving videos, edits weather interactions and light transport separately, and uses constrained video diffusion to improve sensor realism, reaching a 0.2586 CLIP score and 0.915 vehicle detection IoU on Waymo; its main advantage is spatially grounded, parametric weather and lighting control rather than a large numerical lead over generative baselines.

Browse all 52 Autonomous Driving papers →

🤖 Robotics & Embodied AI (20)¶

360CityArena: A Realistic Virtual Urban Navigation Benchmark for Embodied Agents: 360CityArena places 175 manually constructed tasks in an Akihabara street-network pose graph built from real panoramic videos, diagnosing seven urban navigation capabilities; the best evaluated model, Gemini-2.5 Flash, scores only 17.1% overall against 77.3% for locally familiar humans, exposing intertwined weaknesses in map-to-view alignment, sustained exploration, and fine-grained perception.
3DWay: Generalizing Robot Manipulation via 3D Consistent Waypoints: 3DWay expresses 3D trajectory generation as multi-view 2D coordinate prediction in a VLM's text interface, reconstructs executable waypoints using calibrated geometry, reaches 64.0% success on unseen RLBench tasks with fixed top-down control, and improves few-shot pi0 adaptation through local spatial guidance.
AdaDexGrasp: Adaptive Dexterous Grasping via 3D Visuo-Tactile Representation Fusion: AdaDexGrasp predicts which hand part should contact each object region to generate an initial grasp, then maps measured touch back into 3D geometry to assess and correct unstable configurations, raising simulated success on unseen categories from 59% without adaptation to 83%.
AeroVLA: A Vision-Language-Action Model for UAV Navigation via Minimalist End-to-End Control: AeroVLA combines front/down views, coarse target bearings, and native numerical tokens in a reactive OpenVLA-7B policy that jointly controls UAV motion and termination, achieving 37.58% SR on unseen TravelUAV maps versus LongFly's 11.27%, while still relying on a target-bearing prior and a low-level flight controller.
AgentVLN: Towards Agentic Vision-and-Language Navigation: AgentVLN uses Qwen2.5-VL-3B to schedule perception and planning skills through geometrically grounded image prompts, with local correction and on-demand depth queries resolving navigation ambiguities, achieving 67.2% / 69.5% SR on R2R-CE / RxR-CE Val-Unseen.
AMCoNav: Asynchronous Multi-module Collaborative Framework for Embodied Visual Navigation: AMCoNav keeps a lightweight navigation policy running while large models reason on demand in the background, then uses shared context and probability fusion to choose exploration, verification, or direct navigation, achieving 60.0% / 60.0% / 53.3% success on the three HM3D-OVON splits and reducing runtime against a synchronous counterpart without guaranteeing shorter paths.
Attention-DP3: Spatially Object-aware 3D Diffusion Policy via Geometry-aligned Attentional Conditioning: Attention-DP3 projects language-specified object masks onto observed point clouds and conditions an otherwise unchanged DP3 action denoiser with three complementary semantic fields, raising MetaWorld average success from 0.669 to 0.726 and improving real-robot manipulation under clutter.
BeTTER: Diagnose the Illusion of Embodied Reasoning in Vision-Language-Action Models: BeTTER separates motor competence from semantic, compositional and state reasoning through task interventions designed to keep familiar motions executable: three representative VLAs achieve only 5%, 15% and 0% success on unseen subgoal compositions, while physical robot tests also reveal behavioral inertia and phase conflicts.
Beyond Dense Futures: World Models as Structured Planners for Robotic Manipulation: StructVLA trains a world model to predict sparse images selected around gripper transitions and motion pauses, then transfers this task-progress representation to action generation, achieving average success rates of 75.0% on SimplerEnv-WidowX and 94.8% on LIBERO.
Beyond Description: Cognitively Benchmarking Fine-Grained Action for Embodied Agents: CFG-Bench decomposes action understanding into physical interaction, temporal-causal relations, intention, and evaluation, using hybrid QA and counterfactual gating to expose the gap between description and execution knowledge; fine-grained data SFT raises Qwen2.5-VL-7B's average EB-ALFRED result from 4.7 to 9.7.

Browse all 20 Robotics & Embodied AI papers →

📐 Optimization & Theory (1)¶

AnaPFL: When Closed-Form Solutions Meet Generalization and Personalization in Personalized Federated Learning: AnaPFL analytically aggregates a global primary stream and then fits a local residual refinement stream on frozen visual features, requiring one aggregation round and improving accuracy over the strongest baseline in each of 18 benchmark settings by 1.57–16.71 percentage points.

🔗 Causal Inference (1)¶

FD\(^2\): A Dedicated Framework for Fine-Grained Dataset Distillation: Addressing the issue that existing decoupled dataset distillation methods ignore discriminative local regions on fine-grained datasets, FD\(^2\) introduces Counterfactual Attention Learning (CAL) to extract discriminative attention maps and class prototypes. Integrating fine-grained feature constraints and intra-class sample similarity constraints during the distillation phase, it serves as a plug-and-play module that significantly improves the distillation quality of SRe2L++ and FADRM+ on fine-grained datasets such as CUB-200-2011, FGVC-Aircraft, and Stanford Cars, achieving a maximum improvement of +15.1% at IPC=1.

🔬 Interpretability (3)¶

Blind to Position, Biased in Language: Probing Mid-Layer Representational Bias in Vision-Language Encoders for Zero-Shot Language-Grounded Spatial Understanding: B2G probes frozen vision-language encoders and reuses spatially sensitive visual features and more language-consistent text features for region ranking, increasing average multilingual referring segmentation mIoU from 42.42 to 49.62 across nine benchmarks, with additional multilingual inference cost.
Evaluating the Interpretability of Sparse Autoencoders with Concept Annotations: This paper proposes an interpretability evaluation framework for SAEs based on human concept annotations. It features the FBMP algorithm supporting many-to-one matching and the TAPAScore causal validation metric based on target attribute perturbation. Furthermore, two synthetic perturbation datasets, synCUB and synCOCO, are constructed. Experiments show that existing automated evaluation metrics (FMS, MS, CKNNA) fail the sanity checks, whereas the proposed matching metrics and TAPAScore reliably distinguish trained from untrained SAEs. Additionally, increasing overcompleteness is found to degrade the quality of perturbation alignment.
On the Faithfulness of Post-Hoc Concept Bottleneck Models: This paper systematically analyzes the faithfulness of concept projection in post-hoc CBMs. It demonstrates that classifier accuracy cannot serve as a proxy metric for the quality of the bottleneck layer (since random projections can also achieve competitive performance), identifies two fundamental causes of unfaithfulness—covariate shift in auxiliary datasets and systematic errors in VLM surrogate labels—and proposes corresponding diagnostic metrics (\(\mathcal{H}\Delta\mathcal{H}\)-divergence and error-activation correlation).

📦 Model Compression (14)¶

Accurate Zero-shot Quantization via Hierarchical Teacher-Assistant Distillation: Zest connects a full-precision CNN to a low-bit student without original training images through an intermediate-precision teacher assistant, quantization-aware image recalibration, and block-wise reconstruction, reaching 70.50% ImageNet-1K Top-1 accuracy for ResNet-18 W4A4, 0.51 percentage points below the reported full-precision baseline.
Audio-Visual Continual Test-Time Adaptation without Forgetting: AVReCAP retrieves historical fusion-layer parameters using input statistics instead of continually overwriting one parameter state, achieving 43.51% mean accuracy on VGGSound-2C, 6.28 percentage points above the source model, while substantially reducing source-domain forgetting.
AVQ-Attention: Adaptive Vector-Quantized Attention: AVQ-Attention starts with coarse codeword attention and selectively activates learned children for the parents receiving the most attention within each query tile; fused precomputation and incremental correction improve the speed–quality trade-off, reaching 43.33% mIoU in a controlled ADE20K comparison, still below exact attention's 49.0%.
Benchmarking Federated Learning & Knowledge Distillation for Point Cloud Classification: A benchmark of federated teachers and distillation objectives shows that labeled proxy data can produce a 92.94% student from an 8.50% teacher, so a compact student's high accuracy does not necessarily demonstrate successful transfer from federated training.
Boosting Text-Driven Video Segmentation via Geometry-Aware Distillation: Built on SAMWISE, GeoLaV first pretrains on continuous novel views synthesized from single images, then distills semantic and 3D geometric teachers on real videos, reaching 70.5 J&F on Ref-YouTube-VOS—1.3 points above the equally sized 202M-parameter SAMWISE—without retaining teachers or projection heads at inference.
CaPCL: Caption-Preserved Continual Learning for Text-to-Image Retrieval: CaPCL requires the model to continue generating auxiliary captions written for current images prior to the update when learning new text-image retrieval tasks, thereby protecting old semantics not covered by the new query. Across the average scores of four task sequences using BLIP-2, it reduces the forgetting metric (FM) of ModX from 0.0890 to 0.0654, while increasing the average accuracy (AA) from 0.6387 to 0.6942.
Condensing Large-Scale Datasets Directly with Minimal Information Loss: This paper points out that the "data -> model -> image" dual compression process in mainstream large-scale dataset distillation (SRe2L-based family) causes severe information loss and shifts distilled images away from the real distribution, thereby undermining Relabeling. To address this, the authors propose CIM, which uses a computable "effective information gap" metric to directly minimize the information discrepancy between the synthetic and real sets on the original images, bypassing the expensive recovery stage. Consequently, CIM achieves 48.7% Top-1 accuracy (ResNet-18) on ImageNet-1K with IPC=10 in just 80 minutes on a single GPU.
FastSTAR: Spatiotemporal Token Pruning for Efficient Autoregressive Video Synthesis: FastSTAR jointly uses spatial and temporal feature changes to select tokens for computation during InfinityStar's high-resolution refinement, while preventing residual additions at skipped positions, reducing end-to-end 720p text-to-video latency from 81.7 to 40.6 seconds as the VBench total score changes from 83.89 to 83.18.
MambaRaw: Selective State Space Modeling for Efficient 4K Raw Image Reconstruction: MambaRaw introduces State Space Models (SSM/Mamba) into the entropy parameter estimation of the JPEG-guided metadata RAW image reconstruction framework. Through two lightweight modules—energy-guided tile selective scanning (TileMambaBlock) and energy-aware feature refinement (EAR)—the approach simultaneously improves reconstruction quality (PSNR increases by 1.2–1.4 dB) and reduces coding latency (by approximately 9%) at 4K resolution.
MLVC: Multi-Platform Learned Video Codec for Real-World Deployment: By changing the scale parameters required for entropy coding from "computed in real-time by the network" to "deterministically transmitted through the hyperprior," this work enables neural video codecs to perform "encode on A, decode on B" across heterogeneous NPUs (such as Apple, Intel, and Qualcomm) for the first time without crashing. Concurrently, a series of improvements including gated memory, ReGLU, and long-term reference frame recovery are introduced to recoup the rate-distortion loss caused by cross-platform constraints, achieving a >70% BD-rate (MOS) reduction relative to hardware HEVC on video conferencing benchmarks, with an average of 100 FPS across three platforms.

Browse all 14 Model Compression papers →

🏥 Medical Imaging (14)¶

A Dual-space Patch-driven Complementary Learning Framework for Semi-supervised Multi-organ Segmentation: DPCL combines rarity- and uncertainty-guided cropping, full-image–patch pseudo-label fusion, and complementary contrastive objectives to reduce confusion between neighboring organs, reaching 68.89% mean Dice on Synapse with two labeled training cases.
A Mechanism-Driven Theory of Phase Transitions in Active Learning: This paper proposes a mechanism-driven theory of phase transitions in active learning: the four components of the PAC generalization bound (empirical risk, distribution discrepancy, model complexity, confidence term) are reinterpreted as dynamically interacting terms, proving that a shift in their dominance along the labeling trajectory is inevitable. Measurable proxy metrics and piecewise regression are used to identify a "data-driven - transition - model-driven" three-phase taxonomy, explaining why representativeness, coverage, and uncertainty strategies excel at different budget stages.
Active View Selection for Tomography via Perturbed Gaussian Ensemble: To address the challenging problem of active X-ray view selection in sparse-view CT, a perturbed Gaussian ensemble framework based on radiative 3DGS is proposed. By randomly perturbing the density parameters of low-density Gaussian primitives to construct a lightweight model ensemble, epistemic uncertainty is quantified using the structural similarity (SSIM) variance of projections under candidate views. The view that maximizes the exposure of geometric artifacts is selected as the next best acquisition angle.
AlphaRad: Grounded Zero-Shot Classification in Chest Radiology via α-Corrected Binary Cross Entropy and Factorized Latent Supervision: AlphaRad turns chest radiology reports into per-image concept supervision and combines negative-sampled α-corrected binary cross entropy with independently supervised latent subspaces, reaching 83.3% average AUROC across 16 classification datasets at 224 resolution while improving spatial grounding.
Atlas is Your Perfect Context: One-Shot Customization for Generalizable Foundational Medical Image Segmentation: AtlasSegFM converts one annotated anatomical atlas into automatic prompts for a frozen segmentation foundation model and learns voxel-wise fusion at test time, raising SegRap Dice from 50.02 for five-click nnInteractive to 69.19 while retaining a dependence on anatomical correspondence between support and query.
Beyond Isolated Scans: Cross-Phase Alignment of Structure and Topology for 3D Medical Pretraining: CAST adds wavelet-based structural alignment and local feature-relation alignment to reconstruction-based pretraining on paired non-contrast and contrast-enhanced CT, reaching 0.9163 Dice on AIIB23 airway segmentation without outperforming every baseline on every downstream task.
Beyond Random Sampling: Distribution-Aware Alignment for Semi-Supervised Medical Image Segmentation: The framework selects annotation candidates using frozen visual features and density-weighted K-Center, then trains with memory-guided Copy-Paste and progressive branch activation, reaching 87.3% DSC on PROMISE with two labeled cases; the gain combines changed sample selection and changed training.
BrainFIBRE: A Foundation Model via Information Decomposition for Brain Microstructure: BrainFIBRE is the first foundation model for brain tissue microstructure. It treats three NODDI-derived microstructural maps (NDI/ODI/FWF) as three modalities. Using "Self-supervised Partial Information Decomposition (SPID) + Counterfactual Candidate Construction (CCC)", it pre-trains a Mixture-of-Experts (MoE) network on 55,000 subjects from the UK Biobank. It decouples the unique, redundant, and synergistic information of the three modalities without labels, achieving state-of-the-art (SOTA) performance across diverse downstream tasks including age, sex, cerebrovascular and neurodegenerative markers, and cognition in multi-ethnic cohorts. In addition, the expert weights are highly interpretable.
BrainRiem: Riemannian Prototype Learning for Source-Free Cross-Site Brain Network Diagnosis: BrainRiem proposes to learn compact brain network prototypes on the Riemannian manifold via bi-level optimization, achieving cross-site functional brain connectivity diagnosis without access to source data. It substantially outperforms existing methods on the ABIDE and REST-meta-MDD multi-site benchmarks, and the learned prototypes exhibit biological interpretability.
Dual-Prior Guided Null-Space Learning with Mixture-of-Splines for Arbitrary Medical Slice Super-Resolution: This work reformulates arbitrary-scale medical slice super-resolution as a constrained inverse problem. By utilizing range-null space decomposition, it achieves a hard measurement consistency constraint. Within the null space, a Mixture-of-Splines (MoS) adaptively assigns different continuity orders based on local anatomical features, strictly guaranteeing that the acquired slices remain unchanged while generating anatomically plausible interpolated details.

Browse all 14 Medical Imaging papers →

🛡️ AI Safety (9)¶

A Classifier-Agnostic Zero-Shot Adversarial Attack Detection via CLIP: Proposes A4D, a completely black-box, zero-shot adversarial attack detection framework. By exploiting the sensitivity of CLIP to micro-perturbations, the framework compares the cosine similarity of image embeddings with a set of carefully crafted text prompts, and then aggregates them into a single detection score via PCA. This detects adversarial examples without needing to know the attack type or classifier architecture, achieving SOTA performance across multiple attacks, datasets, and classifiers.
Adversarial Attack and Disturbance Detection by Hadamard-Coded Output Representations for Object Detection and Semantic Segmentation: HadamardNet replaces conventional class outputs with redundant codewords and uses probability-simplex projection to produce both class probabilities and an inconsistency residual, enabling single-pass perturbation detection without guaranteeing that predictions on disturbed inputs are correct.
Anti-Prompt: Image Protection against Text-Guided Image-to-Video Generation: Anti-Prompt protects an owner's images with small perturbations that weaken text conditioning, increase the relative dominance of visual pathways, and disrupt image encoding; on white-box CogVideoX evaluation, its video assessment average is 3.25 versus 3.92 for reproduced I2VGuard, where lower scores indicate stronger protection.
AracNet: Revealing Debiasing Signals across Layers with Shallow Monitors: AracNet trains shallow monitors on a frozen biased classifier, selects the layer whose ground-truth confidence distribution best exposes bias, and reweights training of a reinitialized classifier, reaching \(91.27\pm0.56\%\) worst-group accuracy on Waterbirds without relying on a narrowly timed early-stopping point for bias mining.
Exploiting Local Flatness for Efficient Out-of-Distribution Detection: This paper provides the first systematic analysis of the difference in loss landscape curvature between OOD and ID samples, discovering that OOD inputs exhibit larger Hessian curvature which increases as the distribution shift intensifies. Based on this, a lightweight OOD detector, Fold, is proposed, which replaces expensive parameter-space curvature approximations with a feature-space Hessian. This is paired with partial feature normalization to enhance ID-OOD separability, and a self-supervised logit masking scheme (AutoFold) is introduced to automatically calibrate the normalization parameter. Fold achieves an average AUROC improvement of 1.63% and a 2.30% reduction in FPR95 across multiple benchmarks, with a computational overhead comparable to a single forward pass.
Improving Adversarial Robustness via Activation Amplification and Attenuation: This paper proposes A3 (Activation Amplification and Attenuation), a lightweight learnable activation scaling module that implements two modes—activation amplification and attenuation—using the same set of parameters. During training, the degraded predictions from the amplification mode are utilized as negative references to construct contrastive and ranking losses. During inference, only the attenuation mode is employed to enhance adversarial robustness without introducing significant computational overhead.
IREU: Identity-Related Encoder-Only Unlearning for Customized Portrait Generation: IREU introduces the identity unlearning problem for Customized Portrait Generation (CPG) for the first time. By leveraging Face-Swap to localize identity-related dimensions in the embedding space and executing feature perturbations strictly along these dimensions, it establishes an unlearning pipeline that only updates the image encoder to erase target identity generation capabilities while preserving the fidelity of other identities. Furthermore, the unlearned encoder can be transferred to different CPG generators with zero fine-tuning.
Moiré Video Authentication: A Physical Signature Against AI Video Generation: This paper proposes using moiré interference patterns as a physical signature to authenticate video authenticity: a compact double-layer grating structure is placed in the shooting scene, and the Pearson correlation coefficient between the fringe phase change and the translational displacement of the grating image is calculated to distinguish real videos from AI-generated ones. In real videos, these two are strictly coupled by the laws of optical geometry (mean correlation coefficient of 0.87), whereas even the strongest current video generation models (Veo 3.1, Grok Imagine, LTX-2) fail to precisely reproduce this coupling even under optimal configurations (mean correlation coefficient of 0.57, Cohen's d of 1.71, indicating highly significant differences).
ProtoFair: Fair Self-Supervised Contrastive Learning via Pseudo-Counterfactual Pairs: ProtoFair proposes a plug-and-play fairness regularization term. Without modifying existing self-supervised contrastive learning objectives, it identifies "same semantic content but different sensitive groups" pseudo-counterfactual pairs through momentum-updated unsupervised clustering prototypes, pulling these cross-group samples closer in the embedding space to force the encoder to learn representations invariant to sensitive attributes. Combined with SimCLR / SupCon / BarlowTwins / BYOL on CelebA, UTKFace, and NIH Chest X-rays, it significantly reduces Equalized Odds while maintaining competitive accuracy.