PruNeRF: Segment-Centric Dataset Pruning via 3D Spatial Consistency
This addresses the issue of vulnerability to distractors in NeRF for 3D scene learning, which is incremental as it builds on existing methods with refinements.
The paper tackles the problem of distractors in Neural Radiance Fields (NeRF) training by proposing PruNeRF, a segment-centric dataset pruning framework that identifies and removes distractors using 3D spatial consistency, resulting in improved robustness against distractors compared to state-of-the-art methods.
Neural Radiance Fields (NeRF) have shown remarkable performance in learning 3D scenes. However, NeRF exhibits vulnerability when confronted with distractors in the training images -- unexpected objects are present only within specific views, such as moving entities like pedestrians or birds. Excluding distractors during dataset construction is a straightforward solution, but without prior knowledge of their types and quantities, it becomes prohibitively expensive. In this paper, we propose PruNeRF, a segment-centric dataset pruning framework via 3D spatial consistency, that effectively identifies and prunes the distractors. We first examine existing metrics for measuring pixel-wise distraction and introduce Influence Functions for more accurate measurements. Then, we assess 3D spatial consistency using a depth-based reprojection technique to obtain 3D-aware distraction. Furthermore, we incorporate segmentation for pixel-to-segment refinement, enabling more precise identification. Our experiments on benchmark datasets demonstrate that PruNeRF consistently outperforms state-of-the-art methods in robustness against distractors.