Perception and Decision Making Intergration for Autonomous Driving

Our main focus is on researching 2D image perception and multi-sensor 3D surround perception methods. We aim to explore solutions that integrate perception and decision making end-to-end and create a universal framework for autonomous driving. Our goal is to advance autonomous driving technology into a new phase.

Representative Work:

Generic Framwork for Bird’s-Eye-View (BEV) Recognition, Unified BEV Feature and Model Optimization

• BEVFormer: Learning Bird’s-Eye-View Representation from Multi-Camera Images via Spatiotemporal Transformers

  [ECCV 2022 6th most influential paper] [The 100 most cited AI papers of 2022] [Ranking 1st on Waymo 2022]

• BEVFormer v2: Adapting Modern Image Backbones to Bird’s-Eye-View Recognition via Perspective Supervision

  [CVPR 2023 highlight paper]

End-to-end Autonomous Driving Framework with Perception and Decision-making Integration

• Planning-oriented Autonomous Driving

  [CVPR 2023 best paper]