V2X Perception

V2X perception aims to enhance autonomous vehicle perception by collaboratively fusing sensor data from multiple vehicles and roadside units (RSUs) via V2X communication, overcoming limitations of individual vehicle sensors like occlusion and limited range. Current research focuses on developing robust algorithms, often based on transformer architectures, to effectively fuse heterogeneous sensor data (e.g., camera images, LiDAR point clouds) from diverse agents, addressing challenges like asynchronous communication and sensor variations. This research is crucial for improving the safety and reliability of autonomous driving systems, particularly in complex and dynamic environments, by providing a more comprehensive and accurate understanding of the surrounding scene.