Research map / Detection and segmentation / 3D detection and driving
Delving Into the Devils of Bird’s-Eye-View Perception: A Review, Evaluation and Recipe
Hongyang Li, Chonghao Sima, Jifeng Dai, Wenhai Wang, Lewei Lu, Huijie Wang, Jia Zeng, Zhiqi Li, Jiazhi Yang, Hanming Deng, Hao Tian, Enze Xie and 10 more
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 conv offsets in Temporal Self Attention (ablation) | BEVFormer++ (R101, mini) · Waymo val LET-mAPL | LET-mAPL: 34.6 -> 35.9 | Improved |
| add deformable view encoder (ablation) | BEVFormer++ (R101, mini) · Waymo val LET-mAPL | LET-mAPL: 34.6 -> 36.1 | Improved |
| add global location regression (ablation) | BEVFormer++ (R101, mini) · Waymo val | L1/mAPH: 25.5 -> 27.2 | Improved |
| add multi-scale and flip augmentation (ablation) | BEVFormer++ (R101, mini) · Waymo val | L1/mAPH: 25.5 -> 26.8 | Improved |
| add Smooth L1 loss (ablation) | BEVFormer++ (R101, mini) · Waymo val | L1/mAPH: 25.5 -> 27.3 | Improved |
| add learnable loss weight (ablation) | BEVFormer++ (R101, mini) · Waymo val | L1/mAPH: 25.5 -> 26.5 | Improved |
| add sync BN (ablation) | BEVFormer++ (R101, mini) · Waymo val | L1/mAPH: 25.5 -> 25.6 | No clear effect |
| add 2x BEV scale (ablation) | BEVFormer++ (R101, mini) · Waymo val | L1/mAPH: 25.5 -> 25.5 | No clear effect |
| add LET-IoU based assignment (ablation) | BEVFormer++ (R101, mini) · Waymo val | LET-mAPL: 34.6 -> 34.7 | No clear effect |
| add label smoothing (ablation) | BEVFormer++ (R101, mini) · Waymo val | LET-mAPL: 34.6 -> 36.0 | Improved |
| add test-time augmentation (ablation) | Voxel-SPVCNN · Waymo LiDAR seg val | mIoU: 69.6 -> 71.1 | Improved |
| add temporal multi-frame input (ablation) | Voxel-SPVCNN · Waymo LiDAR seg val | mIoU: 71.6 -> 72.4 | Improved |
Closest papers by meaning
- Benchmarking and Improving Bird’s Eye View Perception Robustness in Autonomous Driving (TPAMI 2025 · similarity 0.93)
- Toward Real-world BEV Perception: Depth Uncertainty Estimation via Gaussian Splatting (CVPR 2025 · similarity 0.92)
- Improving Bird's Eye View Semantic Segmentation by Task Decomposition (CVPR 2024 · similarity 0.91)
- M-BEV: Masked BEV Perception for Robust Autonomous Driving (AAAI 2024 · similarity 0.90)
- MamBEV: Enabling State Space Models to Learn Birds-Eye-View Representations (ICLR 2025 · similarity 0.89)
- A Black-Box Evaluation Framework for Semantic Robustness in Bird’s Eye View Detection (AAAI 2025 · similarity 0.89)
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.
