Research map / Detection and segmentation / 3D detection and driving
CoIn3D: Revisiting Configuration-Invariant Multi-Camera 3D Object Detection
Zhaonian Kuang, Rui Ding, Haotian Wang, Xinhu Zheng, Meng Yang, Gang Hua
Read the paperCheck your idea against 137,217 papers
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 SFM and CDA together (ablation) | BEVDepth-R50 · Nuscenes->Waymo | NDS*: 0.178 -> 0.513 | Improved |
| apply CDA alone without SFM (ablation) | BEVDepth-R50 · Nuscenes->Waymo | NDS*: 0.178 -> 0.224 | Mixed |
| add SFM without CDA (ablation) | BEVDepth-R50 · Nuscenes->Waymo | NDS*: 0.178 -> 0.358 | Improved |
| add camera-aware SE module (CA) to SFM+CDA (ablation) | BEVDepth-R50 · Nuscenes->Waymo | NDS*: 0.513 -> 0.504 | Hurt |
| add ground depth map prior (ablation) | SFM · Nuscenes->Waymo | NDS*: 0.462 -> 0.498 | Improved |
| add ground gradient map prior (ablation) | SFM · Nuscenes->Waymo | NDS*: 0.498 -> 0.506 | Improved |
| add Plucker raymap prior (ablation) | SFM · Nuscenes->Waymo | NDS*: 0.506 -> 0.513 | Improved |
| add inverse focal map (ablation) | CDA baseline · Nuscenes->Waymo | NDS*: 0.224 -> 0.462 | Improved |
| add novel view synthesis augmentation to focal augmentation (ablation) | SFM · Nuscenes->Waymo | NDS*: 0.418 -> 0.513 | Improved |
| add focal augmentation (ablation) | SFM · Nuscenes->Waymo | NDS*: 0.358 -> 0.418 | Improved |
| apply CoIn3D to BEVFormer-tiny | BEVFormer-tiny · Nuscenes->Waymo | NDS*: 0.069 -> 0.379 | Improved |
| apply CoIn3D to PETR-vov | PETR-vov · Nuscenes->Lyft | NDS*: 0.046 -> 0.456 | Improved |
Closest papers by meaning
- Towards Generalizable Multi-Camera 3D Object Detection via Perspective Rendering (AAAI 2025 · similarity 0.89)
- Uni3D: A Unified Baseline for Multi-Dataset 3D Object Detection (CVPR 2023 · similarity 0.88)
- LabelAny3D: Label Any Object 3D in the Wild (NeurIPS 2025 · similarity 0.87)
- MonoPlace3D: Learning 3D-Aware Object Placement for 3D Monocular Detection (CVPR 2025 · similarity 0.86)
- ImageNet3D: Towards General-Purpose Object-Level 3D Understanding (NeurIPS 2024 · similarity 0.86)
- Cubify Anything: Scaling Indoor 3D Object Detection (CVPR 2025 · similarity 0.86)
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.
