Skip to content

POp-GS: Next Best View in 3D-Gaussian Splatting with P-Optimality

Conference: CVPR 2025
arXiv: 2503.07819
Code: None
Area: 3D Vision
Keywords: 3D Gaussian Splatting, uncertainty quantification, optimal experimental design, next best view, Fisher information

TL;DR

Introduces the P-Optimality theory from classical optimal experimental design into 3D-GS, deriving a general covariance matrix based on the Hessian matrix. Two approximation schemes, diagonal and block-diagonal, are proposed, significantly outperforming the information gain quantification of FisherRF under D-Optimality and T-Optimality criteria.

Background & Motivation

Although 3D-GS achieves high rendering quality, it lacks inherent uncertainty quantification capabilities, limiting its application in SLAM, active perception, and other fields:

  1. Limitations of FisherRF: Though it quantifies information gain via diagonal approximation of Fisher information, it neglects correlation between parameters and fails to utilize the rich literature of optimal experimental design.
  2. Large Covariance Matrix: 3D-GS may contain millions of parameters, making the full covariance matrix computationally and memory-wise intractable.
  3. Lack of Unified Framework: Existing methods independently design information metrics, lacking a systematic theoretical framework.

This work derives the covariance matrix of 3D-GS from the perspective of maximum likelihood estimation and applies P-Optimality theory to provide a family of information metric solutions.

Method

Overall Architecture

From the perspective of maximum likelihood, the covariance matrix of 3D-GS parameters \(\theta\) is \(\Sigma = \sigma_e^2 (J^TJ)^{-1}\), where \(J\) is the Jacobian of the rendering function with respect to the parameters. After adding the candidate image \(i\), the new Hessian is \(H_i = H_- + J_i^T J_i\). Information metrics are defined through different \(p\) values of P-Optimality.

Key Designs

1. P-Optimality-based Information Quantification Framework

  • Function: Provides a unified family of information gain metrics, where different \(p\) values correspond to different geometric meanings.
  • Mechanism: \(U_p(\Sigma_i) = (\frac{1}{l} \text{trace}(\Sigma_i^p))^{1/p}\). T-Optimality (\(p=1\)) represents average variance (trace), A-Optimality (\(p=-1\)) represents harmonic mean variance, D-Optimality (\(p \to 0\)) represents the volume of the covariance hyper-ellipsoid (determinant), and E-Optimality (\(p \to \pm\infty\)) represents extreme eigenvalues.
  • Design Motivation: D-Optimality possesses monotonicity guarantees in active mapping (uncertainty decreases monotonically with exploration) and corresponds to the differential entropy of multivariate Gaussians from an information-theoretic perspective.

2. Block-Diagonal Covariance Approximation

  • Function: Captures the correlations between parameters of the same ellipsoid on top of the diagonal approximation.
  • Mechanism: Approximates the full Hessian matrix as a block-diagonal matrix, where each block contains all parameters of a single 3D ellipsoid (position, rotation, scale, opacity, color). Channel-wise pixel gradients are calculated to avoid singularity issues, and the block matrices can be processed in parallel on the GPU.
  • Design Motivation: Parameters within the same ellipsoid are most likely to be correlated (e.g., changes in position affect the color contribution), and diagonal approximation completely ignores these correlations.

3. Batch Selection Algorithm

  • Function: Iteratively selects a subset of views with the highest information gains from a set of candidate views.
  • Mechanism: Greedy strategy—iteratively selects the candidate image that maximizes the improvement of the P-Optimality metric, updates the Hessian, and repeats. No extra training is required, and redundancy between views is captured via incremental Hessian updates.
  • Design Motivation: Single-view selection neglects redundancy between views, whereas batch selection naturally handles redundancy by updating the covariance through the Hessian.

Loss & Training

No extra training loss is involved. The information quantification is calculated based on the Hessian of the pre-trained 3D-GS model:

\[H_i = H_- + J_i^T J_i, \quad J = \frac{\partial h}{\partial \theta}\bigg|_{\theta_*, p_i}\]

Key Experimental Results

Blender Dataset (Select 10 views from 100 candidates)

Method PSNR↑ SSIM↑ LPIPS↓
Uniform 25.82 0.944 0.051
FisherRF 27.14 0.956 0.039
Diag T-Opt (Ours) 27.89 0.960 0.035
Block D-Opt (Ours) 28.31 0.963 0.032

Mip-NeRF360 Dataset

Method PSNR↑ SSIM↑ LPIPS↓
Uniform 22.15 0.698 0.271
FisherRF 23.42 0.732 0.243
Block D-Opt (Ours) 24.18 0.756 0.221

Comparison of Different \(p\) Values in P-Optimality

Metric Criterion Approximation Method Blender PSNR↑
T-Optimality (p=1) Diagonal 27.89
D-Optimality (p→0) Diagonal 27.95
T-Optimality (p=1) Block 28.12
D-Optimality (p→0) Block 28.31

Key Findings

  • Block D-Optimality outperforms FisherRF by ~1.2 PSNR on Blender and ~0.8 PSNR on Mip-NeRF360.
  • D-Optimality consistently outperforms T/A/E-Optimality, aligning with theoretical expectations (monotonicity guarantee + information-theoretic significance).
  • Block-diagonal approximation improves over simple diagonal approximation by ~0.4 PSNR, validating the importance of parameter correlations.
  • It does not require candidate image content; only camera poses are needed to evaluate the info gain.

Highlights & Insights

  1. Elegant Theoretical Framework: Unifies 3D-GS information quantification into classical optimal experimental design, providing a class of solutions with theoretical guarantees.
  2. Practical Block-Diagonal Approximation: Effectively captures parameter correlations with acceptable overhead increase.
  3. No Extra Training Required: Purely based on gradient information of the pre-trained model, plug-and-play.

Limitations & Future Work

  • The computational cost of block-diagonal approximation is still high, cubic in the dimension of ellipsoid parameters.
  • Changes in parameter counts during 3D-GS densification/pruning are not considered.
  • Greedy batch selection is not globally optimal; more efficient combinatorial optimization methods can be explored.
  • Extension to active perception in dynamic scenes can be explored in the future.
  • FisherRF: A pioneer in 3D-GS information quantification, but only utilizes diagonal Fisher information.
  • Classical SLAM Literature: P-Optimality is widely applied in keyframe selection and loop closure detection.
  • 3D-GS Pruning: Block-diagonal approximation is also applied to identify redundant ellipsoids.

Rating

⭐⭐⭐⭐ — Represents a solid theoretical contribution, successfully introducing classical optimal design theory to 3D-GS. Experiments consistently outperform baselines across multiple datasets, and the block-diagonal approximation is a practical innovation.