This paper introduce RaCFusion, an attribute-aware radar-camera fusion framework for 3D object detection. It should highlight the core idea of using a camera stream as the main detector and improving it with a Radar-assisted hierarchical refinement process. The description should contrast this approach with traditional symmetrical fusion methods and briefly mention the two key refinement modules: Radar-assisted Query Generation and Radar-assisted Velocity Aggregation.

If you use our method in your research, please cite our paper:

@misc{RaCfusion,
  title={Improving Camera-based 3D Object Detection via Radar-assisted Hierarchical Refinement },
  author= {Yingjie Wang, Jiajun Deng, Yuenan Hou, Yao Li, Lidian Wang, Yanyong Zhang},
  journal= {RAL},
  year = {2026},
}