Research map / Detection and segmentation / 3D detection and driving
CRAFT: Camera-Radar 3D Object Detection with Spatio-Contextual Fusion Transformer
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 camera-radar proposal-level fusion with Soft Polar Association and SCFT | CRAFT-I · nuScenes test | mAP: 33.8 -> 41.1 | Helped |
| Add camera-radar proposal-level fusion with Soft Polar Association and SCFT | CRAFT-I · nuScenes val | mAP: 32.2 -> 41.1 | Helped |
| Use attention-based transformer fusion rather than radar-only fusion (ablation) | CRAFT-I, nuScenes val car · nuScenes val, car | mean AP: 61.8 -> 65.1 | Helped |
| Use attention-based transformer fusion rather than feature concatenation (ablation) | CRAFT-I, nuScenes val car · nuScenes val, car | mean AP: 63.1 -> 65.1 | Helped |
| Use Soft Polar Association instead of RoIPool (ablation) | CRAFT, nuScenes val car · nuScenes val, car | mean AP: 58.7 -> 65.1 | Helped |
| Use Soft Polar Association instead of Ball Query (ablation) | CRAFT, nuScenes val car · nuScenes val, car | mean AP: 64.3 -> 65.1 | Helped |
| Use polar rather than Cartesian coordinates for association and regression (ablation) | CRAFT, nuScenes val car · nuScenes val, car | mean AP: 58.6 -> 65.1 | Helped |
| Add location-offset prediction (ablation) | CRAFT, nuScenes val car · nuScenes val, car | mAP: 53.1 -> 69.6 | Helped |
| Add velocity prediction (ablation) | CRAFT, nuScenes val car · nuScenes val, car | mAP: 65.4 -> 69.6 | Helped |
| Add dimension prediction head (ablation) | CRAFT, nuScenes val car · nuScenes val, car | mAP: 69.6 -> 65.1 | Hurt |
| Add rotation prediction head (ablation) | CRAFT, nuScenes val car · nuScenes val, car | mAP: 69.6 -> 65.7 | Hurt |
Closest papers by meaning
- CRT-Fusion: Camera, Radar, Temporal Fusion Using Motion Information for 3D Object Detection (NeurIPS 2024 · similarity 0.89)
- HGSFusion: Radar-Camera Fusion with Hybrid Generation and Synchronization for 3D Object Detection (AAAI 2025 · similarity 0.89)
- RobuRCDet: Enhancing Robustness of Radar-Camera Fusion in Bird's Eye View for 3D Object Detection (ICLR 2025 · similarity 0.88)
- RICCARDO: Radar Hit Prediction and Convolution for Camera-Radar 3D Object Detection (CVPR 2025 · similarity 0.88)
- Towards Robust 3D Object Detection with LiDAR and 4D Radar Fusion in Various Weather Conditions (CVPR 2024 · similarity 0.88)
- RCBEVDet: Radar-camera Fusion in Bird's Eye View for 3D Object Detection (CVPR 2024 · similarity 0.87)
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.