Ming-Yang Cheng
National Cheng Kung University, Kao Yuan University, Beihang University
Papers
16
Total Citations
168
H-Index
8
About
Ming-Yang Cheng is a leading figure in intelligent robotics and automation, whose work bridges the gap between machine vision, control theory, and human-robot collaboration. His research centers on developing autonomous systems capable of complex manipulation and safe physical interaction. A cornerstone of his contribution is the integration of deep reinforcement learning with computer vision for robotic grasping, as demonstrated in his highly cited 2023 work (37 citations), which enables robots to self-learn and adapt to small-volume, large-variety production tasks. Cheng has also made seminal advances in bipedal locomotion on uneven terrain (19 citations), moving beyond idealized flat surfaces to address real-world challenges. His expertise extends to visual servoing, where he has pioneered PH-spline and Kalman filter-based methods for precise contour tracking and depth estimation. Notably, his recent work on closed-loop input error identification (2024, 14 citations) and observer-based force-sensorless control (2020, 13 citations) provides efficient, safe solutions for collaborative robots. With over 150 total citations across his top papers, Cheng’s research is instrumental in creating more capable, intuitive, and industrially viable robotic systems.
Research Focus
Key Achievements
Top Papers
- 1Vision-Based Robotic Object Grasping—A Deep Reinforcement Learning Approach37 citations · 2023
- 2
- 3Dynamic biped robot locomotion on less structured surfaces19 citations · 2000
- 4
- 5
- 6
- 7
- 8
- 9Vision-based 3-D object pick-and-place tasks of industrial manipulator7 citations · 2017
- 10