My research interests are mainly in the field of computer vision and robotics, including but not limited to object detection, classification, unsupervised learning, 3D scene understanding, etc.

At present, I’m a researcher at MEGVII(Face++) Research, focusing on computer vision related tasks such as 2D/3D object detection, and unsupervised learning. I’m mentored by Zeming Li and Dr. Gang Yu, under the supervision of of Dr. Jian Sun. More information can be found at CV.

I got my B.E. from Software Engineering College (Distinguished Engineer Program) of South China University of Technology (one of the “Project 985” of China) with the First-Class Honor in July 2018.

The most recent AutoAssign proposes a novel label assignment paradigm for object detection and achieves SOTA(52.1% AP) on MS COCO. More information releated to research can be found at Publications or Google Scholar.

I build the world’s first 3D object detection framework Det3D, and won the 1st place of nuScenes 3D object detection challenge (CVPR 2019). At Megvii Research, I write an efficient and general computer vision toolkit named cvpack2 (cvpods). It supports many computer vision tasks, and is now widely used at MEGVII Research.


Conference Reviewer: CVPR’21