Research map / Detection and segmentation / 3D detection and driving
CODA: A Real-World Road Corner Case Dataset for Object Detection in Autonomous Driving
Kaican Li, Kai Chen, Haoyu Wang, Lanqing Hong, Chaoqiang Ye, Jianhua Han, Yukuai Chen, Wei Zhang, Chunjing Xu, Dit-Yan Yeung, Xiaodan Liang, Zhenguo Li and 1 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 |
|---|---|---|---|
| raise background-removal overlap ratio from 0.15 to 0.45 (ablation) | COPG · CODA-ONCE proposals | AR: 1.1 -> 5.1 | Improved |
| raise background-removal overlap ratio from 0.45 to 0.75 (ablation) | COPG · CODA-ONCE proposals | AR: 5.1 -> 6.1; proposals 1522 -> 2315 | Mixed |
| raise minimum clustering angle from 4 to 8 degrees (ablation) | COPG · CODA-ONCE proposals | AR: 3.1 -> 5.1 | Improved |
| raise minimum clustering angle from 8 to 16 degrees (ablation) | COPG · CODA-ONCE proposals | AR: 5.1 -> 4.7 | No clear effect |
| raise cluster size threshold from 5 to 20 points (ablation) | COPG · CODA-ONCE proposals | AR: 5.1 -> 5.1 | No clear effect |
| raise max distance from 25 to 50 (ablation) | COPG · CODA-ONCE proposals | AR: 4.5 -> 5.1 | Improved |
| raise common-class suppression IoU threshold from 0.00 to 0.25 (ablation) | COPG · CODA-ONCE proposals | AR: 2.7 -> 5.1 | Improved |
| use COPG instead of Faster R-CNN to detect corner cases | COPG · CODA-KITTI | AR50 large: 8.3 -> 44.9 | Improved |
| use Cascade R-CNN instead of Faster R-CNN | SODA10M detectors · CODA CORNER | AR*: 6.8 -> 8.3 | Improved |
| use open-world ORE instead of Faster R-CNN | SODA10M detectors · CODA CORNER | AR*: 6.8 -> 8.3 | Improved |
| pretrain on BDD100K instead of SODA10M | RetinaNet · CODA CORNER | AR*: 11.9 -> 12.8 | Improved |
| use Cascade R-CNN instead of RetinaNet | SODA10M detectors · CODA CORNER | AR*: 11.9 -> 8.3 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 4 on most shared results.
- Faster R-CNN — Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks (TPAMI 2016) · better on 3 of 4 shared results
- Sparse R-CNN — Sparse R-CNN: End-to-End Object Detection With Learnable Proposals (CVPR 2021) · better on 1 of 3 shared results
- Deformable DETR — Deformable DETR: Deformable Transformers for End-to-End Object Detection (ICLR 2021) · better on 2 of 3 shared results
- Cascade R-CNN — Cascade R-CNN: Delving Into High Quality Object Detection (CVPR 2018) · better on 1 of 3 shared results
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
- 3D Object Detection From Images for Autonomous Driving: A Survey (TPAMI 2023 · similarity 0.90)
- One Million Scenes for Autonomous Driving: ONCE Dataset (NeurIPS 2021 · similarity 0.90)
- OpenAD: Open-World Autonomous Driving Benchmark for 3D Object Detection (NeurIPS 2025 · similarity 0.89)
- SDAC: A Multimodal Synthetic Dataset for Anomaly and Corner Case Detection in Autonomous Driving (AAAI 2024 · similarity 0.88)
- CorrBEV: Multi-View 3D Object Detection by Correlation Learning with Multi-modal Prototypes (CVPR 2025 · similarity 0.88)
- AIDE: An Automatic Data Engine for Object Detection in Autonomous Driving (CVPR 2024 · 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.
