Yinhuai Wang

Hi! I am a first-year Ph.D student at HKUST, under the supervision of Prof. Ping Tan. I hold a Master's degree from Peking University and a Bachelor's degree from Xidian University. Additionally, I have interned at IDEA Research and Unitree Robotics.

Before my postgraduate studies, I worked at DH-Robotics as one of the initial team members (the first engineer), where I led a team in the development of electronics and algorithms for robot grippers and arms.

Email / CV / Google Scholar / Github / Zhihu

Research

My current research interest lies in the intersection of Computer Vision, Machine Learning, and Robotics. My long-term goal is to enable robots to master all human skills. Below are some selected papers.

	SkillMimic: Learning Reusable Basketball Skills from Demonstrations Yinhuai Wang, Qihan Zhao, Runyi Yu, Ailing Zeng, Jing Lin, Zhengyi Luo, Hok Wai Tsui, Jiwen Yu, Xiu Li, Qifeng Chen, Jian Zhang, Lei Zhang, Ping Tan CVPR*, 2025 project page / arXiv / code We for the first time enable simulated humanoids to learn reusable basketball skills purely from demonstrations. The learned skills can be reused for complex and long-term tasks such as continuous scoring.
	PhysHOI: Physics-Based Imitation of Dynamic Human-Object Interaction Yinhuai Wang, Jing Lin, Ailing Zeng, Zhengyi Luo, Jian Zhang, Lei Zhang arXiv, 2023 project page / arXiv / code We enable physically simulated humanoids to imitate interactions from video demonstrations, without designing task-specific rewards.
	Zero-Shot Image Restoration Using Denoising Diffusion Null-Space Model Yinhuai Wang, Jiwen Yu, Jian Zhang ICLR, 2023 (Oral Presentation) project page / arXiv / code We bring Range-Null space Decomposition (RND) into diffusion models, enabling diverse image restoration tasks in a zero-shot manner, without extra training or optimization.
	GAN Prior based Null-Space Learning for Consistent Super-Resolution Yinhuai Wang, Yujie Hu, Jiwen Yu, Jian Zhang AAAI, 2023 (Oral Presentation) code / arXiv We bring Range-Null space Decomposition (RND) into GAN-Prior based SR models to accelerate the convergence and ensure the downsampling consistency.
	Freedom: Training-free energy-guided conditional diffusion model Jiwen Yu, Yinhuai Wang, Chen Zhao, Bernard Ghanem, Jian Zhang ICCV, 2023 code / arXiv / FreeDoM is a simple but effective training-free method generating results under control from various conditions using unconditional diffusion models.
	LaPE: Layer-adaptive Position Embedding for Vision Transformers with Independent Layer Normalization Runyi Yu, Zhennan Wang, Yinhuai Wang, Kehan Li, Chang Liu, Haoyi Duan, Xiangyang Ji, Jie Chen, ICCV*, 2023 code / arXiv We find that simply adding an independent LN to each layer can robustly improve the performance of vision transformers.

Misc.

	Travel around the world from 2019.7 to 2020.1 - Riding bicycle through Xinjiang, Tibet, Nipel, and India - Footprints span Germany, Malaysia, Nipel, India, UAE, Iran, Turkey, Lebanon, Egypt, Saudi Arabia, Ethiopia, Kenya, Tanzania, Rwanda, Hongkong, Macao, Taiwan, and Mainland China.
	Build a Robot Arm From Scratch - I built a two-axis robot arm with self-designed motor driver, FK & IK algorithm, trajectory generation, and 2D impedance control. 2018~2019 Zhihu
	One of the original creators of these cool grippers, 2017~2018 I did - The PCB design. - FOC Motor Control algorithm. - Online Trajectory Generation algorithm. - Force Control and Impedance Control algorithm.
	Reviewer for CVPR, ICCV, ICLR, NeurIPS, SIGGRAPH, AAAI, TIP, IJCV, and TPAMI

This cool template is stolen from Jon Barron!