Ye Shi 石野

Ye Shi 石野

Assistant Professor, 研究员, 博导, ShanghaiTech University

Dr. Ye Shi received the Ph.D. degree at the University of Technology Sydney, Australia in 2018. Dr Shi served as a Research Assistant at the University of New South Wales, Australia and a Postdoctoral Fellow at the University of Technology Sydney from 2017 to 2020. Dr. Shi has been an Assistant Professor (PI) in the School of Information Science and Technology at ShanghaiTech University since January 2021. He is leading the YesAI Lab at ShanghaiTech University.

✡️ Research Interests

Our recent research interests mainly focus on the Embodied AI, Generative Models, Large Multimodal Models, and on the fundamental optimizations underlying them. We are interested in developing AI algorithms that are more controllable, robust, efficient, and safe. Moreover, our target is to use the developed AI methods for various scientific applications, e.g., humanoid robots, smart grids, etc.

🔊 Position Openings

  • PhD Students: 1-2 PhD positions available with topics of Generative Models and Embodied AI.
  • Postgraduate Students: Dr. Ye Shi recruits 2~3 postgraduate students for each academic year. The candidate is expected to be self-motivated in one of the following areas: Generative Models, Reinforcement Learning, Embodied AI, Large Multimodal Models, Humanoid Robots, Smart Grids, etc.
  • Undergraduates/Visting Students: We warmly welcome students majoring in Computer Science, Mathematics, Robotics, Electrical Engineering and other related disciplines to join our group.
  • Research Assistant: Dr. Ye Shi is seeking a research assistant to work closely with the principal investigator and students in the laboratory. A Bachelor’s or master’s degree in mathematics, computer science, Robotics, electrical engineering, control, or related areas is required.

A solid mathematical background and sufficient programming skills are required. If you are interested in this opening, please email me your CV.

🔥 News

  • 2026.7: Our paper "Hierarchical Constrained Reinforcement Learning with Dynamic Boundary for Spatio-Temporal Vehicle-to-Grid Scheduling" received the Best Student Paper Award Runner-Up at World Artificial Intelligence Conference Academic 2026. Congratulations to Haoyu Yan, Shutong Ding, Jiebao Zhang, Xi Yao, Yu Liu, Haoyu Wang, and Chenchi Luo!
  • 2026.7: ACM MM 2026 +3. 100% Acceptance Rate.
  • 2026.5: ICML 2026 +7.
  • 2026.4: I will serve as an Area Chair for CoRL 2026.
  • 2026.3: I gave a talk at Nvidia GTC 2026.
  • 2025.9: NeurIPS 2025 +3. One is Oral (77 among 21575).
  • 2025.9: I gave a talk at CSIG 2025 at Qingdao.
  • 2025.9: I will serve as an Area Chair for ICLR 2026.
  • 2025.8: I gave a talk at AI for Engineering Conference 2025 at Harbin.
  • 2025.7: Our paper was accepted by ICCV 2025 as a Highlight.
  • 2025.6: We will host a Workshop "Human-Robot-Scene Interaction and Collaboration" at ICCV 2025.
  • 2025.6: I gave a talk "Diffusion-Driven Embodied AI: Breakthroughs in Theory and Algorithmic Frontiers" at VALSE 2025.
  • 2025.5: We hosted a Workshop "Humanoid Locomotion and Interaction" at the 2025 Zhangjiang Embodied AI Conference.
  • 2025.5: One paper was accepted by ICML 2025 as a Spotlight. Congratulations to Kaizhen Zhu and Mokai Pan! 2 Submissions with 1 Acceptance.
  • 2025.4: Our paper UniDB won Outstanding Paper Award of ICLR 2025 DeLTa Workshop. Congratulations to Kaizhen Zhu and Mokai Pan!
  • 2025.4: I will serve as an Area Chair for NeurIPS 2025.
  • 2025.2: Three Papers accepted by CVPR 2025. Congratulations to Bikang Pan, Qun Li (CVPR Highlight); Shijie Wu, Yihang Zhu; Chunlin Yu, Hanqing Wang. 4 Submissions with 3 Acceptances.
  • 2024.10: I gave a talk at RLChina 2024: Controllable diffusion and optimization models for generation and decision in Embodied AI.
  • 2024.9: Three Papers accepted by NeurIPS 2024. Congratulations to Shutong Ding, Bikang Pan, Haixiang Sun. 3 Submissions with 3 Acceptances.
  • 2024.7: One Paper accepted by ACM Multimedia 2024. Congratulations to Chaofan Huo, 1 Submission with 1 Acceptance.
  • 2024.5: Two Papers accepted by ICML 2024. Congratulations to Lingxiao Yang and Tianyu Cui. 3 Submissions with 2 Acceptances.
  • 2024.3: I gave a talk at The 1st China Embodied AI Conference: "Model-Data Hybrid-Driven Embodied Agents: Fundamentals and Applications".
  • 2024.3: Our Paper "A Distributionally Robust Model Predictive Control for Static and Dynamic Uncertainties in Smart Grids" has been accepted by IEEE Trans. Smart Grid. Congratulations to Qi Li!
  • 2024.2: Three Papers accepted by CVPR 2024. 4 Submissions with 3 Acceptances.
  • 2024.1: Our Paper "Understanding Convergence and Generalization in Federated Learning through Feature Learning Theory" has been accepted by ICLR 2024. 1 Submission with 1 Acceptance.
  • 2023.12: Two Papers accepted by AAAI 2024. 2 Submissions with 2 Acceptances.
  • 2023.11: I gave a talk at RLChina 2023: "Towards Responsible Decision and Control via Implicit Networks".
  • 2023.09: Five Papers accepted by NeurIPS 2023. 6 Submissions with 5 Acceptances.

📝 Selected Publications

arXiv 2026
sym

SteerGenPO: Steering Generative Reinforcement Learning into Stable Robotic Controller, arXiv 2026.

Yixuan Wang#, Shutong Ding#, Ke Hu#, Tianxiang Gui, Jingya Wang, Ye Shi*

[paper]

ICML 2026
sym

CReL: Conformal Reliability: A New Evaluation Metric for Conditional Generation, ICML 2026.

Yachen Gao, Xinwei Sun, Yikai Wang, Ye Shi, Jingya Wang, Jianfeng Feng, Yanwei Fu

[paper] [project]

ICML 2026
sym

CGPO: Sample-Efficient Diffusion-based Reinforcement Learning with Critic Guidance, ICML 2026.

Shutong Ding#, Zejia Zhong#, Zhongyi Wang, Ke Hu, Bikang Pan, Jingya Wang, Ye Shi*

[paper] [code] [project]

ICML 2026
sym

DiscoForcing: A Unified Framework for Real-Time Audio-Driven Character Control with Diffusion Forcing, ICML 2026.

Kaiyang Ji#, Bingsheng Qian#, Binghuan Wu, Kangyi Chen, Ye Shi, Jingya Wang*

[paper] [project] [code]

ICML 2026
sym

DBC: Distributional Reinforcement Learning with Diffusion Bridge Critics, ICML 2026.

Shutong Ding#, Yimiao Zhou#, Ke Hu, Mokai Pan, Shan Zhong, Yanwei Fu, Jingya Wang, Ye Shi*

[paper]

ICML 2026
sym

Sample from What You See: Visuomotor Policy Learning via Diffusion Bridge with Observation-Embedded Stochastic Differential Equation, ICML 2026.

Zhaoyang Liu, Mokai Pan, Zhongyi Wang, Kaizhen Zhu, Haotao Lu, Jingya Wang, Ye Shi*

[paper] [project] [code]

ICML 2026
sym

Diffusion Bridge or Flow Matching? A Unifying Framework and Comparative Analysis, ICML 2026.

Kaizhen Zhu#, Mokai Pan#, Zhechuan Yu, Jingya Wang, Jingyi Yu, Ye Shi*

[paper] [code]

ICML 2026
sym

DiOpt: Diffusion-based Learning Framework for Constrained Nonconvex Optimization with Weighted Bootstrapped Refinement, ICML 2026.

Shutong Ding#, Yimiao Zhou#, Ke Hu, Xi Yao, Junchi Yan, Xiaoying Tang, Ye Shi*

[paper] [code] [project]

ICLR 2026
sym

Human-Object Interaction via Automatically Designed VLM-Guided Motion Policy, ICLR 2026.

Zekai Deng, Ye Shi, Kaiyang Ji, Lan Xu, Shaoli Huang, Jingya Wang*

[paper] [project] [code]

IEEE TPAMI 2026
sym

UniDB++: A Unified and Fast-Sampling Diffusion Bridge Framework via Stochastic Optimal Control, IEEE Transactions on Pattern Analysis and Machine Intelligence, 2026.

Mokai Pan#, Kaizhen Zhu#, Yuexin Ma, Yanwei Fu, Jingyi Yu, Jingya Wang, Ye Shi*

[paper] [code]

arXiv 2025
sym

DreamPolicy: A Unified World-model Policy for Scalable Humanoid Locomotion, arXiv 2025.

Yahao Fan#, Tianxiang Gui#, Kaiyang Ji#, Shutong Ding, Chixuan Zhang, Yifeng Xu, Ke Yang, Jiayuan Gu, Jingyi Yu, Jingya Wang*, Ye Shi*

[paper] [project]

NeurIPS 2025 Oral
sym

OpenHOI: Open-World Hand-Object Interaction Synthesis with Multimodal Large Language Model, NeurIPS 2025 Oral (0.3%).

Zhenhao Zhang, Ye Shi* , Lingxiao Yang, Suting Ni, Qi Ye, Jingya Wang*

[paper]

NeurIPS 2025
sym

GenPO: Generative Diffusion Models Meet On-Policy Reinforcement Learning, NeurIPS 2025.

Shutong Ding#, Ke Hu#, Shan Zhong, Haoyang Luo, Weinan Zhang, Jingya Wang, Jun Wang, Ye Shi*

[paper] [code] [project]

ICCV 2025 Highlight
sym

Towards Immersive Human-X Interaction: A Real-Time Framework for Physically Plausible Motion Synthesis, ICCV 2025 Highlight.

Kaiyang Ji, Ye Shi*, Zichen Jin, Kangyi Chen, Lan Xu, Yuexin Ma, Jingyi Yu, Jingya Wang*

[paper] [project]

ICML 2025 Spotlight
sym

UniDB: A Unified Diffusion Bridge Framework via Stochastic Optimal Control, ICML 2025 Spotlight.

Kaizhen Zhu#, Mokai Pan#, Yuexin Ma, Yanwei Fu, Jingyi Yu, Jingya Wang, Ye Shi*

[paper] [project] [code]

CVPR 2025 Highlight
sym

NLPrompt: Noise-Label Prompt Learning for Vision-Language Models, CVPR 2025 Highlight.

Bikang Pan#, Qun Li#, Xiaoying Tang, Wei Huang, Zhen Fang, Feng Liu, Jingya Wang, Jingyi Yu, Ye Shi*

[paper] [code]

ArXiv
sym

Exploring the Boundary of Diffusion-based Methods for Solving Constrained Optimization, arXiv:2502.10330

Shutong Ding, Yimiao Zhou, Ke Hu, Xi Yao, Junchi Yan, Xiaoying Tang, Ye Shi*

[paper]

ArXiv
sym

Human-Object Interaction via Automatically Designed VLM-Guided Motion Policy, ArXiv:2503.18349

Zekai Deng, Ye Shi, Kaiyang Ji, Lan Xu, Shaoli Huang, Jingya Wang

[paper] [project]

CVPR 2025
sym

AffordDP: Generalizable Diffusion Policy with Transferable Affordance, CVPR 2025.

Shijie Wu#, Yihang Zhu#, Yunao Huang, Kaizhen Zhu, Jiayuan Gu, Jingyi Yu, Ye Shi*, Jingya Wang*

[paper] [project] [code]

CVPR 2025
sym

SeqAfford: Sequential 3D Affordance Reasoning via Multimodal Large Language Model, CVPR 2025.

Chunlin Yu#, Hanqing Wang#, Ye Shi, Haoyang Luo, Sibei Yang, Jingyi Yu, Jingya Wang

[paper] [project] [code]

NeurIPS 2024
sym

Diffusion-based Reinforcement Learning via Q-weighted Variational Policy Optimization, NeurIPS 2024

Shutong Ding, Ke Hu, Zhenhao Zhang, Kan Ren, Weinan Zhang, Jingyi Yu, Jingya Wang, Ye Shi*

[paper] [project] [code]

NeurIPS 2024
sym

Federated Learning from Vision-Language Foundation Models: Theoretical Analysis and Method, NeurIPS 2024

Bikang Pan, Wei Huang*, Ye Shi*

[paper] [code]

ICML 2024
sym

Guidance with Spherical Gaussian Constraint for Conditional Diffusion, ICML 2024

Lingxiao Yang, Shutong Ding, Yifan Cai, Jingyi Yu, Jingya Wang, Ye Shi*

[paper] [code]

ICML 2024
sym

Harmonizing Generalization and Personalization in Federated Prompt Learning, ICML 2024

Tianyu Cui, Hongxia Li, Jingya Wang, Ye Shi*

[paper] [code]

CVPR 2024
sym

Global and Local Prompts Cooperation via Optimal Transport for Federated Learning, CVPR 2024

Hongxia Li, Wei Huang, Jingya Wang, Ye Shi*

[paper] [code]

CVPR 2024
sym

S2Fusion: A Unified Diffusion Framework for Scene-aware Human Motion Estimation from Sparse Signals, CVPR 2024

Jiangnan Tang, Jingya Wang*, Kaiyang Ji, Lan Xu, Jingyi Yu, Ye Shi*

[paper] [code]

NeurIPS 2023
sym

Reduced Policy Optimization for Continuous Control with Hard Constraints, NeurIPS 2023

Shutong Ding, Jingya Wang, Yali Du, Ye Shi*

[paper] [code]

NeurIPS 2023
sym

Two Sides of The Same Coin: Bridging Deep Equilibrium Models and Neural ODEs via Homotopy Continuation, NeurIPS 2023

Shutong Ding#, Tianyu Cui#, Jingya Wang, Ye Shi*

[paper] [code]

NeurIPS 2023
sym

CSOT: Curriculum and Structure-Aware Optimal Transport for Learning with Noisy Labels, NeurIPS 2023

Wanxing Chang, Ye Shi, Jingya Wang

[paper] [project] [code]

NeurIPS 2023
sym

Fed-CO2: Cooperation of Online and Offline Models for Severe Data Heterogeneity in Federated Learning, NeurIPS 2023

Zhongyi Cai, Ye Shi*, Wei Huang, Jingya Wang

[paper] [code]

NeurIPS 2023
sym

Contextually Affinitive Neighborhood Refinery for Deep Clustering, NeurIPS 2023

Chunlin Yu, Ye Shi, Jingya Wang

[paper] [code]

ICCV 2023
sym

Knowledge-Aware Federated Active Learning with Non-IID Data, ICCV, 2023

Yu-Tong Cao, Ye Shi, Jingya Wang, Baosheng Yu, Dacheng Tao

[paper] [code]

IEEE TNNLS 2023
sym

FedTP: Federated Learning by Transformer Personalization, IEEE Transactions on Neural Networks and Learning Systems, 2023

Hongxia Li#, Zhongyi Cai#, Jingya Wang, Jiangnan Tang, Weipng Ding, Chin-Teng Lin, Ye Shi*

[paper] [code]

ICLR 2023
sym

Alternating Differentiation for Optimization Layers, ICLR 2023

Haixiang Sun, Ye Shi*, Jingya Wang, Hoang Duong Tuan, H.V. Poor, Dacheng Tao

[paper] [code]

AAAI 2023 (oral)
sym

IKOL: Inverse kinematics optimization layer for 3D human pose and shape estimation via Gauss-Newton differentiation, AAAI 2023 (oral)

Juze Zhang, Ye Shi*, Yuexin Ma, Lan Xu, Jingyi Yu, Jingya Wang*

[paper] [project] [code]

AAAI 2023 (oral)
sym

Lifelong Person Re-Identification via Knowledge Refreshing and Consolidation, AAAI 2023 (oral)

Chunlin Yu, Ye Shi, Zimo Liu, Shenghua Gao, Jingya Wang*

[paper] [project] [code]

NeurIPS 2022 (Spotlight)
sym

Unified Optimal Transport Framework for Universal Domain Adaptation, NeurIPS 2022 (Spotlight)

Wanxing Chang, Ye Shi*, Hoang Duong Tuan, Jingya Wang*

[paper] [project] [code] [video] [VALSE]

📖 Educations

  • 2014.02 - 2018.11, Ph.D., University of Technology Sydney, NSW, Australia.
  • 2009.09 - 2013.06, B.S., Northwestern Polytechnical University, Xi’an, China.

💻 Work Experience

  • Tenure-track Assistant Professor (2021.01 -now): School of Information Science and Technology, ShanghaiTech University.
  • Postdoctoral Fellow (2019.07-2020.12):
    • Australian Artificial Intelligence Institute, School of Computer Science, University of Technology Sydney.
    • Supervisor: Prof. Chin-Teng Lin (IEEE Fellow), E-mail: chin-teng.lin@uts.edu.au.
  • Research Assistant (2018.11-2019.06):
    • School of Electrical Engineering and Telecommunications, Faculty of Engineering, University of New South Wales.
    • Supervisor: Prof. Andrey V. Savkin, E-mail: a.savkin@unsw.edu.au

🎖️ Awards

  • 2025 Our paper UniDB won Outstanding Paper Award of ICLR 2025 DeLTa Workshop.
  • 2024 My students received the NeurIPS 2023 Scholar Award.
  • 2021 Best Student Paper Award (Corresponding author) at Australia Artificial Intelligence Institute.
  • 2019 Outstanding Overseas Students Award, Chinese Ministry of Education, Australia. (A total of 500 people worldwide while only 50 people in Australia).
  • 2018 FEIT PhD Post Thesis Publication Award, University of Technology Sydney 2018, Australia.
  • 2017 Higher Degree Research Publication Award, University of Technology Sydney, Australia.
  • 2016 Best Paper Award, the 6th IEEE International Conference on Control Systems, Computing and Engineering, Malaysia.
  • 2016 ARC Discovery Scholarship, the University of Technology Sydney, 2014-2016, Australia.
  • 2016 International Research Scholarships, University of Technology Sydney, 2014-2016, Australia.
  • 2013 Meritorious Winner of the Interdisciplinary Contest in Modeling (ICM), The Society for Industrial and Applied Mathematics, America.
  • 2012 First Prize of Chinese Undergraduate Mathematical Contest for Modeling (CUMCM), China Society for Industrial and Applied Mathematics, China.
  • 2010 National Scholarship, Chinese Ministry of Education, China.

💬 Talks

  • 2025.06, Diffusion-driven Embodied AI: Breakthroughs in Theory and Algorithmic Frontiers, the Vision And Learning Seminar (VALSE) 2025.
  • 2025.05, Generative Models for Humanoid Locomotion and Manipulation, the 2025 Zhangjiang Embodied AI Conference.
  • 2024.10, Controllable diffusion and optimization models for generation and decision in Embodied AI, RLChina 2024.
  • 2024.04, Model-Data Hybrid-Driven Embodied Agents: Fundamentals and Applications, The 1st China Embodied AI Conference.
  • 2023.11, Towards Responsible Decision and Control via Implicit Networks, RLChina 2023.
  • 2019.10, Joint Coordination of Plug-in Electrical Vehicles Charging and Smart Grid Operations, IEEE International Conference on Communications, Control, and Computing Technologies for Smart Grids, Beijing.
  • 2018.04, Nonconvex and nonsmooth optimization in smart grids, International Forum for Interdisciplinary Sciences and Engineering Open Forum, Wuhan University.
  • 2016.11, Nonconvex Spectral Algorithm for Solving BMI on the Reduced Order H∞ Control, IEEE International Conference on Control Systems, Computing and Engineering, Penang, Malaysia.

🕴️ Activities

  • Area Chair NeurIPS 2025, ICLR 2026.
  • Session Chair The 10th IEEE International Conference on Communications, Control, and Computing Technologies for Smart Grids (SmartGridComm 2019).
  • Program committee
    • NeurIPS 2023, AAAI 2023, AAAI 2024, CVPR 2024, ICLR 2024, ICML 2024, NeurIPS 2024, AAAI 2025, ICLR 2025, etc.
  • Reviewer
    • TNNLS, TIP, JSAC, TSG, TFS, TSE, TII, TIE, TIV, TIM, TMLR, IJRNC, etc.

🧑‍🏫 Teaching

  • Course SI251 - Convex Optimization, ShanghaiTech University, 2021 Spring, 2021 Autumn, 2022 Autumn, 2024 Autumn.
  • Course SI152 - Numerical Optimization, ShanghaiTech University, 2022 Spring, 2023 Spring, 2024 Spring.