Binghui Zuo

Binghui Zuo | 左炳辉

I'm a Ph.D. student at Southeast University. Starting from 2021, I have studied in Vision and Cognition Lab (VCL), advised by Prof. Yangang Wang. Prior to that, I got my B.E. degree (Top 1% graduates) of automation at Qingdao University of Technology in 2021, supervised by Prof. Zhao.

I'm interested in 3D computer vision and computer graphics. Much of my research is about reconstructing single hand, interacting hands or hand-object from images. Currently, I am attempting to generate personalized human avatars from monocular images with diffusion models.

Email / Google Scholar / Twitter / Github

News

[2024.09] 🎉🎉🎉 One paper accepted by TVCG!!! My first top journal as the first author!
[2024.02] One paper accepted by SPL.
[2023.07] Two papers accepted by ICCV2023!!! My first top conference as the first author!
[2023.03] One paper accepted by CVPR2023.
[2023.01] One paper accepted by TVCG2023, and another accepted by GM2023.
[2022.03] One paper accepted by CVPR 2022.
[2021.12] Create this Homepage.

Publications

	GraspDiff: Grasping Generation for Hand-Object Interaction With Multimodal Guided Diffusion Binghui Zuo, Zimeng Zhao, Wenqian Sun, Xiaohan Yuan, Zhipeng Yu, Yangang Wang IEEE Transactions on Visualization and Computer Graphics (TVCG), 2024 Paper
	Music Conditioned Generation for Human-centric Video Zimeng Zhao, Binghui Zuo, Yangang Wang IEEE Signal Processing Letters (SPL), 2024 Paper Our key idea is to build a temporal generation framework dominated by DDPM and assisted by VAE and GAN. To produce videos with both long duration and high quality, our framework first generates small-scale keyframes and then generates high-resolution videos.
	Reconstructing Interacting Hands with Interaction Prior from Monocular Images Binghui Zuo, Zimeng Zhao, Wenqian Sun, Wei Xie, Zhou Xue, Yangang Wang IEEE/CVF International Conference on Computer Vision (ICCV), 2023 Paper / Project / Code [comming soon] The key idea is first to construct a two-hand interaction prior and recast the interaction reconstruction task as the conditional sampling from the prior.
	Nonrigid Object Contact Estimation With Regional Unwrapping Transformer Wei Xie, Zimeng Zhao, Shiying Li, Binghui Zuo, Yangang Wang IEEE/CVF International Conference on Computer Vision (ICCV), 2023 Paper / Project A novel hand-object contact representation called RUPs (Region Unwrapping Profiles) is proposed, which unwrap the roughly estimated hand-object surfaces as multiple high-resolution 2D regional profiles.
	Semi-supervised Hand Appearance Recovery via Structure Disentanglement and Dual Adversarial Discrimination Zimeng Zhao, Binghui Zuo, Zhiyu Long, Yangang Wang IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023 Paper / Project A solution is proposed to recover clear hand appearance from degraded images. In a semi-supervised learning manner, we first disentangle the bare hand structure from those degraded images and then wrap the appearance to this structure with a dual adversarial discrimination (DAD) scheme.
	Skeleton Extraction for Articulated Objects with the Spherical Unwrapping Profiles Zimeng Zhao, Wei Xie, Binghui Zuo, Yangang Wang IEEE Transactions on Visualization and Computer Graphics (TVCG), 2023 Paper / Project A novel unwrapping method named Spherical Unwrapping Profiles is proposed, which maps a surface into image planes independent of mesh topologies.
	Stability-driven Contact Reconstruction From Monocular Color Images Zimeng Zhao, Binghui Zuo, Wei Xie, Yangang Wang IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022 Paper / Project We use physical engine to optimize reconstructed hand-object interaction, thus produce more realistic result.
	HMDO : Markerless Multi-view Hand Manipulation Capture with Deformable Objects Wei Xie, Zhipeng Yu, Zimeng Zhao, Binghui Zuo, Yangang Wang International Conference on Computational Visual Media (CVM), 2023 Paper / Project We construct the first markerless deformable interaction dataset recording interactive motions of the hands and deformable object. Besides that, a baseline for hand manipulating with deformable objects is proposed.
	Implicit Representation for Interacting Hands Reconstruction from Monocular Color Images Binghui Zuo, Zimeng Zhao, Wei Xie, Yangang Wang International Conference on Image and Graphics (ICIG, Oral), 2023 Paper A novel framework for representing the surface of interacting hands with implicit functions is proposed, which implicitly reconstructs interacting hands with image features as query strategy and condition.