Yongtao Ge | 葛涌涛

I am a Ph.D candidate in Computer Science, working in AIML, the University of Adelaide, with Prof. Chunhua Shen. Before that, I received the M. Eng degree from Southeast University under the supervision by Prof. Wankou Yang. I got my B. Eng degree in Control Science and Engineering from Nanjing University of Science and Technology. I also spent wonderful time at Sensetime, Damo Academy and Tecent.

Currently, my research interest mainly lies in computer vision, including 2D human pose estimation, 3D human reconstruction, human motion generation, and multi-modality scene understanding.

I will graduate in the late 2024 and I am looking for industry positions or Post-doc opportunities. Please feel free to reach out (yongtao.ge@adelaide.edu.au) if you have openings and find me might be a fit.


Mar 29, 2024 Release HumanWild, feel free to try our Huggingface Demo. 🎉
Jul 18, 2023 Zolly is accepted by ICCV 2023, selected as oral (top 1.8%). :sparkles:
Mar 27, 2023 Release Zolly, focusing on perspective-distorted 3D human pose and shape estimation. :sparkles:
Nov 19, 2022 One paper is accepted by AAAI 2023. :sparkles:
Jul 10, 2022 One paper is accepted by ECCV 2022. :sparkles:


  1. humanwild.gif
    3D Human Reconstruction in the Wild with Synthetic Data Using Generative Models
    Yongtao GeWenjia Wang, Yongfan Chen, Hao Chen, and Chunhua Shen
    In arXiv.org, 2024
  2. genpercept_pipeline.png
    Diffusion Models Trained with Large Data Are Transferable Visual Models
    Guangkai Xu, Yongtao Ge, Mingyu Liu, Chengxiang Fan, Kangyang Xie, Zhiyue Zhao, Hao Chen, and Chunhua Shen
    In arXiv.org, 2024
  3. zolly.png
    Zolly: Zoom Focal Length Correctly for Perspective-Distorted Human Mesh Reconstruction
    Wenjia WangYongtao Ge, Haiyi Mei, Zhongang Cai, Qingping Sun, Yanjun Wang, Chunhua ShenLei Yang, and Komura Taku
    In Proc. of the IEEE International Conf. on Computer Vision (ICCV Oral), 2023
  4. poseur_arch.jpg
    Poseur: Direct Human Pose Regression with Transformers
    In Proc. of the European Conf. on Computer Vision (ECCV), 2022
  5. point_teaching_arch.jpg
    Point-Teaching: Weakly Semi-Supervised Object Detection with Point Annotations