Xingyu Chen (陈星宇)



I am currently a Postdoctoral Research Fellow at Peking University.

Before joining PKU, I have 3.5 years of experience in AI corporations, Kuaishou and Xiaobing, working closely with Baoyaun Wang.

I obtained my Ph.D. in Control Science and Engineering from Institute of Automation, Chinese Academy of Science in 2020, under the supervision of Prof. Junzhi Yu. Previously I received my B.S. in Electrical Engineering and Automation from Chengdu University of Technology in 2015.

My research interests lie in human-centric modeling, understanding, interaction, and intelligence.

profile photo
profile photo
Recent Publications

(here for full publications)

HAVE-FUN: Human Avatar Reconstruction from Few-Shot Unconstrained Images
Xihe Yang*, Xingyu Chen *, Shaohui Wang, Daiheng Gao, Xiaoguang Han, Baoyuan Wang (* for equal contribution, for corresponding author)
2024 IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2024.
[PDF] [Project] [BibTeX]

Our HaveFun framework can create human avatars from few-shot unconstrained images.

Mimic3D: Thriving 3D-Aware GANs via 3D-to-2D Imitation
Xingyu Chen*, Yu Deng*, Baoyuan Wang (* for equal contribution)
2023 IEEE International Conference on Computer Vision, ICCV 2023.
[PDF] [Project] [BibTeX]

We propose a novel 3D-aware GAN that can directly render high-resolution images.

Hand Avatar: Free-Pose Hand Animation and Rendering from Monocular Video
Xingyu Chen, Baoyuan Wang, Heung-Yeung Shum
2023 IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2023.
[PDF] [Project] [BibTeX]

We propose a HandAvatar framework, the first method for neural hand rendering with self-occluded illumination.

MobRecon: Mobile-Friendly Hand Mesh Reconstruction from Monocular Image
Xingyu Chen, Yufeng Liu, Yajiao Dong, Xiong Zhang, Chongyang Ma, Yanmin Xiong, Yuan Zhang, Xiaoyan Guo
2022 IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2022.
[PDF] [Code] [BibTeX]

We propose a framework for single-view hand mesh reconstruction, which can simultaneously achieve high reconstruction accuracy, fast inference speed, and temporal coherence.

Camera-Space Hand Mesh Recovery via Semantic Aggregation and Adaptive 2D-1D Registration
Xingyu Chen, Yufeng Liu, Chongyang Ma, Jianlong Chang, Huayan Wang, Tian Chen, Xiaoyan Guo, Pengfei Wan, Wen Zheng
2021 IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2021.
[PDF] [Code] [BibTeX]

We reconstruct 3D hand by dividing camera-space mesh recovery into two sub-tasks, i.e., root-relative mesh recovery and root recovery.

Joint Anchor-Feature Refinement for Real-Time Accurate Object Detection in Images and Videos
Xingyu Chen, Junzhi Yu, Shihan Kong, Zhengxing Wu, and Li Wen
IEEE Transactions on Circuits and Systems for Video Technology (TCSVT), 2021.
[PDF] [BibTeX]

We propose a temporal detection method based on anchor and feature offset refinements.

Towards real-time advancement of underwater visual quality with GAN
Xingyu Chen, Junzhi Yu, Shihan Kong, Zhengxing Wu, Xi Fang, and Li Wen
IEEE Transactions on Industrial Electronics (TIE), 2019.
[PDF] [BibTeX]

We propose a GAN-RS to elevate underwater visual quality.

The website template was adapted from Yu Deng.