Zheng Ge

Researcher at Megvii Inc., Beijing, China.

prof_pic.jpg

He currentlty works as a researcher at Megvii Technology, supervised by Dr.Xiangyu Zhang. He leads a team working on unified foundation models for autonomous driving, as well as 2D detection models on solving real-world challenges. His team develops and maintains several well-known projects, including YOLOX and BEVDepth series.

He obtained his Ph.D degree in Computer Science at Waseda University, supervised by Prof.Osamu Yoshie in 2022. He got his Bachelor degree in ECE department of Southeast University in 2017.

He is also interested in Multi-modal Foundation Models, AIGC and Self-supervised Representation Learning.

news

Jan 1, 2023 One paper is accepted by ICLR 2023.
Nov 1, 2022 Two papers are accepted by AAAI 2023.
Sep 1, 2022 Our team release BEVStereo, a new SOTA camera-only 3D detector on nuScenes dataset.
Jul 1, 2022 One paper is accepted by ECCV 2022.
Apr 1, 2022 I join Megvii Team Base Detection as a researcher, supervised by Zeming Li.
Jul 1, 2021 We release YOLOX , a high performance and anchor free YOLO, with many devices supported.

selected publications

  1. AAAI’23
    BEVStereo: Enhancing Depth Estimation in Multi-view 3d Object Detection with Dynamic Temporal Stereo
    Yinhao Li, Han Bao, Zheng Ge, Jinrong Yang, Jianjian Sun, and 1 more author
    In Proceeding of Association for the Advancement of Artificial Intelligence (AAAI), 2023
  2. AAAI’23
    BEVDepth: Acquisition of Reliable Depth for Multi-view 3d Object Detection
    Yinhao Li, Zheng Ge, Guanyi Yu, Jinrong Yang, Zengran Wang, and 3 more authors
    In Proceeding of Association for the Advancement of Artificial Intelligence (AAAI), 2023
  3. ECCV’22
    Dense Teacher: Dense Pseudo-labels for Semi-supervised Object Detection
    Hongyu Zhou, Zheng Ge, Songtao Liu, Weixin Mao, Zeming Li, and 2 more authors
    In Proceeding of the European Conference on Computer Vision (ECCV), 2022
  4. Arxiv
    YOLOX: Exceeding Yolo Series in 2021
    Zheng Ge, Songtao Liu, Feng Wang, Zeming Li, and Jian Sun
    2021
  5. CVPR’21
    OTA: Optimal Transport Assignment for Object Detection
    Zheng Ge, Songtao Liu, Zeming Li, Osamu Yoshie, and Jian Sun
    In Proceeding of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2021
  6. CVPR’20
    NMS by Representative Region: Towards Crowded Pedestrian Detection by Proposal Pairing
    Xin Huang, Zheng Ge, Zequn Jie, and Osamu Yoshie
    In Proceeding of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2020

Collaborators