Collaborative Perception for Autonomous Driving: Current Status and Future Trend

Autonomous driving is the key technology of intelligent transportation system, and also a very promising engineering project that could fundamentally change the life of human society. Although there has been a lot of progress in both academia and industry in the past decades, autonomous driving is still an important research topic nowadays, especially inspired by the development of computer vision and deep learning recently. One of the crucial module of autonomous driving is perception, which targets to perceiving the surrounding environment and extracting information related to navigation, including object detection, tracking, semantic segmentation and so on. Perception used to be regarded as the technique bottleneck of autonomous driving. In the past few years, perception performance has been significantly improved with increasing large-scale training data and developments of deep learning algorithms. However, it is not enough to meet the demand of practical high level of autonomous capacity because high level or fully autonomous vehicles do not require human supervision and totally rely on the perception of vehicles.