Research map / 3D vision / 3D generation
Viewpoint Equivariance for Multi-View 3D Object Detection
Dian Chen, Jie Li, Vitor Guizilini, Rares Andrei Ambrus, Adrien Gaidon
This paper tried a common change: Add positional encoding
add Fourier+MLP geometric positional encoding and object queries (replacing PETR's encodings) · NuScenes val · mAP: +1.6% mAP (0.404 to 0.420) Improved
What 646 accepted papers report for this change
- 525 improved
- 58 no clear effect
- 7 mixed
- 56 hurt
- Learning-Augmented Scalable Linear Assignment Problem Optimization via Neural Dual Warm-Starts ICML 2026 · add positional encodings to features (d=13 -> d=21) · Block-structured, N=4096 · speedupImproved
- Mesh Based Simulations with Spatial and Temporal awareness ICML 2026 · use learned positional embedding, learnable relative bias or weighted adjacency instead of RoPE · Aneurysm · RMSEHurt
- Sparse ActionGen: Accelerating Diffusion Policy with Real-time Pruning ICML 2026 · add sinusoidal positional encoding to the pruner · Robomimic Tool Hang · success rate (%): 28 -> 54Improved
See this change in detailMatched by the wording of each paper's own result, read by AI from the full text. Papers report gains more often than failures, so this shows what gets published, not the chance a change works.
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add Fourier+MLP geometric positional encoding and object queries (replacing PETR's encodings) (ablation) | PETR-style single-frame detector · NuScenes val | mAP: +1.6% mAP (0.404 to 0.420) | Improved |
| add 2 virtual query views with viewpoint equivariance (joint matching and loss) (ablation) | VEDet single-frame with geometric PE · NuScenes val | mAP: +1.4% mAP (0.420 to 0.434) | Improved |
| add 2 virtual query views with viewpoint equivariance (ablation) | VEDet 2-frame · NuScenes val | mAP: +1.9% mAP (0.432 to 0.451) | Improved |
| add 2-frame temporal modeling (ablation) | VEDet with geometric PE only · NuScenes val | mAP: +1.2% mAP (0.420 to 0.432) | Improved |
| add the extra queries and duplicated box targets without viewpoint equivariance (+2M qry, no VE) (ablation) | VEDet V=0 · NuScenes val | mAP | No clear effect |
| remove Fourier encoding before the MLP (ablation) | VEDet · NuScenes val | mAP: -6.9% mAP (0.451 to 0.382) | Hurt |
| remove camera translation from geometric encoding (ablation) | VEDet · NuScenes val | mAP: -9.2% mAP (0.451 to 0.353) | Hurt |
| remove camera rotation from geometric encoding (ablation) | VEDet · NuScenes val | mAP: -1.5% mAP (0.451 to 0.436) | Hurt |
| treat multi-view predictions of same query as individual objects (no joint matching) (ablation) | VEDet · NuScenes val | mAP: -2.6% mAP (0.451 to 0.425) | Hurt |
| number of virtual query views V increased from 0 to 2 (ablation) | VEDet · NuScenes val | mAP: 0.432 to 0.451 mAP | Improved |
| number of virtual query views increased to 6 (ablation) | VEDet · NuScenes val | mAP: 0.451 to 0.437 mAP | Hurt |
| VEDet vs PETR and PETRv2 at Res-50/Res-101 and V2-99 settings | VEDet · NuScenes val | mAP: +2.0% mAP over PETRv2 (0.451 vs 0.431) | Improved |
Closest papers by meaning
- Towards Generalizable Multi-Camera 3D Object Detection via Perspective Rendering (AAAI 2025 · similarity 0.88)
- Multiview Equivariance Improves 3D Correspondence Understanding with Minimal Feature Finetuning (ICLR 2025 · similarity 0.87)
- MVSDet: Multi-View Indoor 3D Object Detection via Efficient Plane Sweeps (NeurIPS 2024 · similarity 0.85)
- BEVNeXt: Reviving Dense BEV Frameworks for 3D Object Detection (CVPR 2024 · similarity 0.85)
- VGGT: Visual Geometry Grounded Transformer (CVPR 2025 · similarity 0.85)
- GeoBEV: Learning Geometric BEV Representation for Multi-view 3D Object Detection (AAAI 2025 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
