Keypoints-Based Deep Feature Fusion for Cooperative Vehicle Detection of Autonomous Driving

Engineering notes

Compared to a bird's-eye view (BEV) keypoints feature fusion, FPV-RCNN achieves improved detection accuracy by about 9% at a high evaluation criterion (IoU 0.7) on the synthetic dataset COMAP dedicated to collective perception. Moreover, our method also significantly decreases the CPM size to less than 0.3KB, which is about 50 times smaller than the BEV feature map sharing used in previous works.

Chinese explanation / 中文解读

中文解读待补充：本站会优先为端到端自动驾驶、BEV感知、3D目标检测、轨迹预测、路径规划、LiDAR感知等高价值论文补充中文说明。

Original abstract

Sharing collective perception messages (CPM) between vehicles is investigated to decrease occlusions, so as to improve perception accuracy and safety of autonomous driving. However, highly accurate data sharing and low communication overhead is a big challenge for collective perception, especially when real-time communication is required among connected and automated vehicles. In this paper, we propose an efficient and effective keypoints-based deep feature fusion framework built on the 3D object detector PV-RCNN, called Fusion PV-RCNN, (FPV-RCNN for short), for collective perception. We introduce a bounding box proposal matching module of high performance and a keypoints selection strategy to compress the CPM size and solve the multi-vehicle data fusion problem. Besides, we also proposed an effective localization error correction module with maximum consensus to increase the robustness of the data fusion. Compared to a bird's-eye view (BEV) keypoints feature fusion, FPV-RCNN achieves improved detection accuracy by about 9% at a high evaluation criterion (IoU 0.7) on the synthetic dataset COMAP dedicated to collective perception. In addition, its performance is comparable to two raw data fusion baselines that have no data loss in sharing. Moreover, our method also significantly decreases the CPM size to less than 0.3KB, which is about 50 times smaller than the BEV feature map sharing used in previous works. Even with further decreased CPM feature channels, i.e., from 128 to 32, the detection performance does not show obvious drops. The code of our method is available at https://github.com/YuanYunshuang/FPV_RCNN.

Links and sources

• Official / arXiv page
• PDF from original source

Comments