Jingyu Liu

About

I study safe robot learning and trustworthy embodied AI: how robots can anticipate risk, adapt their actions, and operate reliably under uncertainty.

I am a Ph.D. student at Southeast University, advised by Prof. Changyin Sun. Since September 2026, I have been at MBZUAI, working with Asst. Prof. Anqing Duan and Asst. Prof. Xingxing Zuo. Previously, I was a joint doctoral researcher with Prof. Shengbo Eben Li's team at Tsinghua University.

Research

Safe robot learning
Action rectification and safety filters for constrained continuous control.
World-model risk prediction
Diffusion world models and Langevin action search under uncertainty.
Trustworthy embodied AI
Runtime safety for vision-language-action models and robot policies.
Adaptive safety certification
Safety certificates and invariant sets for uncertain robot dynamics.

Publications

Google Scholar

2026

  1. SMART: Sliding Manifold Action Rectification Technique for Safe Reinforcement Learning

    Jingyu Liu, Yuanda Wang, Anqing Duan, Zewen He, Yoshihiko Nakamura, and Changyin Sun

    IEEE Transactions on Industrial Informatics · Early Access, 2026

  2. Safe Reinforcement Learning with Langevin Dynamics for Robotics

    Jingyu Liu, Yuanda Wang, Chenguang Yang, Changyin Sun, and Anqing Duan

    Advanced Engineering Informatics · 76:104974, 2026

  3. M3imic: Learning a Versatile Whole-Body Controller for Multimodal Motion Mimicking

    Zuxing Lu, Ziang Zheng, Yao Lyu, Jingyu Liu, Feihong Zhang, Song Lu, Xin Yuan, Changyin Sun, Xingxing Zuo, and Shengbo Eben Li

    IEEE Robotics and Automation Letters · 11(8):9843-9850, 2026

  4. Safe Online Reinforcement Learning with Diffusion Langevin Policy

    Jingyu Liu, Yuanda Wang, Changyin Sun, and Anqing Duan

    Expert Systems with Applications · Article 131278, 2026

  5. GSFF-SLAM: 3D Semantic Gaussian Splatting SLAM via Feature Field

    Zuxing Lu, Xin Yuan, Shaowen Yang, Jingyu Liu, and Changyin Sun

    Expert Systems with Applications · Article 129601, 2026

2025

  1. Unmanned Surface Vehicle Autonomous Racing and Obstacle Avoidance with Robust Adversarial Deep Reinforcement Learning

    Jingyu Liu, Yuanda Wang, and Changyin Sun

    Engineering Applications of Artificial Intelligence · 161:112250, 2025

Under review

  1. World-Action Models for Robot Learning and Control: A Survey

    Zuxing Lu, Hongjia Zhai, Guanzhi Wang, Huajian Zeng, Jiaqi Yang, Jingyu Liu, Lei Cheng, Yuantai Zhang, Yuheng Qiu, Zezhou Cheng, Ivan Laptev, Danfei Xu, Benjamin Riviere, Giuseppe Loianno, Eric Xing, and Xingxing Zuo

    IEEE Transactions on Robotics · arXiv:2609.16074, 2026

  2. Risk-Sliding Layer for Safe Reinforcement Learning

    Jingyu Liu, Zuxing Lu, Yuanda Wang, and Changyin Sun

    IEEE Transactions on Industrial Electronics

  3. Frontier: Online Adaptive Safety Certification Under Set-Valued Uncertainty

    Jingyu Liu, Yuanda Wang, and Changyin Sun

    Neurocomputing

  4. Unilateral Inequality-Aware Safety Filtering for Mixed Equality-Inequality Constraints in Robot Learning

    Jingyu Liu, Yuanda Wang, Yujie Yang, Yao Lv, Shenbo Li, and Changyin Sun

    IEEE Transactions on Robotics

Experience

  1. Sep. 2026 - Present

    Mohamed bin Zayed University of Artificial Intelligence

    Division of Computing and Mathematical Sciences

    Advisors: Asst. Prof. Anqing Duan and Asst. Prof. Xingxing Zuo

  2. Oct. 2025 - Sep. 2026

    Tsinghua University

    Joint Ph.D. researcher · School of Vehicle and Mobility

    Advisor: Prof. Shengbo Eben Li

  3. Mar. - Aug. 2025

    Mohamed bin Zayed University of Artificial Intelligence

    Visiting student · Robotics Department

    Advisor: Asst. Prof. Anqing Duan

  4. 2022 - 2023

    Huawei Technologies

    General software development engineer · Nanjing

Education

  1. 2023 - Present

    Southeast University

    Ph.D. in Electronic Information

  2. 2019 - 2022

    Nanjing University of Posts and Telecommunications

    M.S. in Control Science and Engineering

  3. 2015 - 2019

    Nanjing University of Posts and Telecommunications

    B.S. in Automation

Get in touch

For research discussions and collaboration, reach me at Jingyu.Liu@seu.edu.cn.

GitHub Bilibili RedNote