- Rating
- 2
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
This paper proposes LiDAR-Anchored Splat-and-Distill for adapting vision foundation encoders to autonomous-driving geometry. During training, LiDAR points are converted into Gaussian surfels, painted with frozen teacher features from context cameras, and rendered into a held-out target camera. A coverage-masked anchor loss constrains regions without rendered supervision, while an auxiliary LiDAR depth head provides metric supervision. LoRA updates are merged into the backbone, so inference remains a standard single-image forward pass without LiDAR.
Critical strengths
- Replacing fragile short-baseline stereo geometry with measured LiDAR is a practical extension of splat-and-distill to surround-view driving rigs, with no added inference cost.
- The paper clearly diagnoses sparse supervision: 76.4% of the target view receives no rendered signal, motivating a simple coverage-masked anchor loss.
Critical weaknesses
- The evidence for the core 3D-aware distillation claim is weak. Distillation alone degrades depth and correspondence; depth gains mainly come from direct depth supervision, while the anchor restores correspondence. Final r@10 gains are only 0.27/0.64 points, with no improvement in the largest-rotation bin.
- Baseline coverage is incomplete and uncontrolled. SnD, FiT3D, and MEF use mismatched training domains, while two highly related works LiDAR-Anchored Collaborative Distillation (arXiv:2602.12524) and DDMS (arXiv:2608.23850) are not discussed. The former particularly weakens the claim that prior LiDAR-to-camera distillation is limited to task-specific BEV models.
- The method is only evaluated on one dataset, which seems insufficient. More datasets such as nuScenes and Waymo are required.
Questions for authors
- I wonder what is contributed by cross-view surfel distillation, given that distillation alone degrades geometry, depth-only gives substantially better depth, and the anchor accounts for most correspondence recovery?
- Why does the method not improve the 10–90° rotation bin and fall below the off-the-shelf encoder beyond 5 m translation? Should the claim of handling arbitrary viewpoint changes be weakened?
- The ablation studies are insufficient. It's encouraged to compare controlled alternatives on the same driving corpus and backbone, including direct LiDAR-to-image feature distillation and measured-geometry SnD without the proposed surfel design.
- Please add discussion about the baselines mentioned above.
- Could you evaluate the method on other public driving datasets (e.g., waymo, nuscenes) or any downstream tasks such as detection or occupancy?
- How sensitive are the results to LiDAR density, vertical field of view, and the uncovered fraction? Can a controlled sparse-LiDAR experiment be provided?