Research map / Detection and segmentation / 3D detection and driving
Exploring Surround-View Fisheye Camera 3D Object Detection
Changcai Li, Wenwei Lin, Zuoxun Hou, Gang Chen, Wei Zhang, Huihui Zhou, Weishi Zheng
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 |
|---|---|---|---|
| transfer pinhole BEVDet to rectified fisheye images (4 fisheye vs 6 pinhole) | BEVDet · Fisheye3DOD test | FDS: 0.563 -> 0.453 | Hurt |
| transfer pinhole PETR to rectified fisheye images | PETR · Fisheye3DOD test | FDS: 0.553 -> 0.411 | Hurt |
| replace perspective rectification with spherical equirectangular feature warping | BEVDet · Fisheye3DOD test | FDS: 0.440 -> 0.485 | Improved |
| replace perspective rectification with spherical equirectangular feature warping | PETR · Fisheye3DOD test | FDS: 0.408 -> 0.470 | Improved |
| use equirectangular instead of cylindrical representation (ablation) | FisheyeBEVDet · Fisheye3DOD test | FDS: 0.476 -> 0.485 | Improved |
| use equirectangular instead of cylindrical representation (ablation) | FisheyePETR · Fisheye3DOD test | FDS: 0.441 -> 0.470 | Improved |
| fisheye spherical modeling instead of pinhole under front-rear camera failure | BEVDet / FisheyeBEVDet · Fisheye3DOD test (sensor failure) | FDS: 0.370 -> 0.454 | Improved |
| fisheye spherical modeling instead of pinhole under front-rear camera failure | PETR / FisheyePETR · Fisheye3DOD test (sensor failure) | FDS: 0.321 -> 0.421 | Improved |
| front-rear layout instead of left-right layout (2 fisheye) (ablation) | FisheyePETR · Fisheye3DOD test | FDS: 0.382 -> 0.421 | Improved |
| front-rear layout instead of left-right layout (2 fisheye) (ablation) | FisheyeBEVDet · Fisheye3DOD test | FDS: 0.431 -> 0.454 | Improved |
| full 4-fisheye surround instead of 2 front-rear (ablation) | FisheyeBEVDet · Fisheye3DOD test | FDS: 0.454 -> 0.485 | Improved |
| full 4-fisheye surround instead of 2 front-rear (ablation) | FisheyePETR · Fisheye3DOD test | FDS: 0.421 -> 0.470 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.
- PETR — PETR: Position Embedding Transformation for Multi-View 3D Object Detection (ECCV 2022) · better on 0 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- GeoBEV: Learning Geometric BEV Representation for Multi-view 3D Object Detection (AAAI 2025 · similarity 0.85)
- Towards Generalizable Multi-Camera 3D Object Detection via Perspective Rendering (AAAI 2025 · similarity 0.85)
- Far3D: Expanding the Horizon for Surround-View 3D Object Detection (AAAI 2024 · similarity 0.84)
- UniMODE: Unified Monocular 3D Object Detection (CVPR 2024 · similarity 0.84)
- Instance-Aware Multi-Camera 3D Object Detection with Structural Priors Mining and Self-Boosting Learning (AAAI 2024 · similarity 0.83)
- RobuRCDet: Enhancing Robustness of Radar-Camera Fusion in Bird's Eye View for 3D Object Detection (ICLR 2025 · similarity 0.83)
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.
