RayFusion: Ray Fusion Enhanced Collaborative Visual Perception
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909832654618624 |
|---|---|
| author | Wang, Shaohong Lu, Bin Xiao, Xinyu Zhong, Hanzhi Pang, Bowen Wang, Tong Xiang, Zhiyu Shan, Hangguan Liu, Eryun |
| author_facet | Wang, Shaohong Lu, Bin Xiao, Xinyu Zhong, Hanzhi Pang, Bowen Wang, Tong Xiang, Zhiyu Shan, Hangguan Liu, Eryun |
| contents | Collaborative visual perception methods have gained widespread attention in the autonomous driving community in recent years due to their ability to address sensor limitation problems. However, the absence of explicit depth information often makes it difficult for camera-based perception systems, e.g., 3D object detection, to generate accurate predictions. To alleviate the ambiguity in depth estimation, we propose RayFusion, a ray-based fusion method for collaborative visual perception. Using ray occupancy information from collaborators, RayFusion reduces redundancy and false positive predictions along camera rays, enhancing the detection performance of purely camera-based collaborative perception systems. Comprehensive experiments show that our method consistently outperforms existing state-of-the-art models, substantially advancing the performance of collaborative visual perception. The code is available at https://github.com/wangsh0111/RayFusion. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_08017 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | RayFusion: Ray Fusion Enhanced Collaborative Visual Perception Wang, Shaohong Lu, Bin Xiao, Xinyu Zhong, Hanzhi Pang, Bowen Wang, Tong Xiang, Zhiyu Shan, Hangguan Liu, Eryun Computer Vision and Pattern Recognition Collaborative visual perception methods have gained widespread attention in the autonomous driving community in recent years due to their ability to address sensor limitation problems. However, the absence of explicit depth information often makes it difficult for camera-based perception systems, e.g., 3D object detection, to generate accurate predictions. To alleviate the ambiguity in depth estimation, we propose RayFusion, a ray-based fusion method for collaborative visual perception. Using ray occupancy information from collaborators, RayFusion reduces redundancy and false positive predictions along camera rays, enhancing the detection performance of purely camera-based collaborative perception systems. Comprehensive experiments show that our method consistently outperforms existing state-of-the-art models, substantially advancing the performance of collaborative visual perception. The code is available at https://github.com/wangsh0111/RayFusion. |
| title | RayFusion: Ray Fusion Enhanced Collaborative Visual Perception |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2510.08017 |