About

Yunduan Cui is a robotics and artificial intelligence researcher whose work sits at the intersection of reinforcement learning, robot control, and human-robot interaction. His research focuses primarily on developing stable, sample-efficient reinforcement learning algorithms and applying them to challenging robotic manipulation tasks, including the notoriously complex domain of cloth handling. Cui's most influential contribution — a deep reinforcement learning framework with smooth policy updates for robotic cloth manipulation (179 citations) — demonstrated that high-dimensional visual inputs could be effectively leveraged for dexterous robot learning. He has consistently addressed core limitations of reinforcement learning, tackling instability and data inefficiency through innovations such as kernel dynamic policy programming, relative entropy regularization, and contact-safe model-based approaches. His 2023 work on continuous dynamic policy programming reflects his ongoing commitment to principled, theoretically grounded algorithm design, extended further into multi-agent systems in 2024. Beyond manipulation, Cui has explored pneumatic artificial muscle control, quaternion neural networks for inverse kinematics, and adaptive human-robot motor skill learning, showcasing remarkable breadth. With over 380 cumulative citations, his contributions are shaping how robots learn safely and efficiently from experience — a foundation increasingly critical to deploying intelligent robots in real-world environments.

Research Focus

Key Achievements

11
H-Index
23
Papers
473
Total Citations
21
Avg Citations/Paper
🏆 Most Cited Paper
Deep reinforcement learning with smooth policy update: Application to robotic cloth manipulation
179 citations · 2018
📈 Most Prolific Year: 2018 (3 Papers)
🤝 Key Collaborators: 35
🏛 Institutions: Nara Institute of Science and Technology, Shenzhen Institutes of Advanced Technology, Doshisha University, Graphic Era University, Chinese Academy of Sciences

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 13 days ago