Papers
20
Total Citations
160
H-Index
7
About
Mingmin Liu is a robotics researcher whose work spans human-robot interaction, autonomous navigation, manipulation, and legged locomotion — areas at the forefront of making robots safer and more capable in real-world environments. Liu's early contributions focused on quadrupedal robotics, with foundational work on centroidal momentum dynamics and foot trajectory planning that established methods for stable, dynamic locomotion. This groundwork evolved into a broader research agenda addressing the critical challenge of robot safety, most notably through highly cited work on real-time collision avoidance planning for human-robot interaction (34 citations) and risk-aware deep reinforcement learning for crowd navigation (21 citations), demonstrating Liu's commitment to deploying robots responsibly alongside humans. In manipulation, Liu advanced robotic grasping through the HTC-Grasp hybrid Transformer-CNN architecture (21 citations) and, more recently, vision-language-action modeling for task-oriented grasping, reflecting an embrace of large multimodal AI. Additional contributions in 3D LiDAR SLAM, trajectory optimization with jerk constraints, and online motion planning round out a versatile portfolio. With over 140 cumulative citations, Liu's research consistently bridges theoretical rigor with practical deployment, making it essential reading for those working on intelligent, collaborative robotic systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2HTC-Grasp: A Hybrid Transformer-CNN Architecture for Robotic Grasp Detection21 citations · 2023
- 3Risk-Aware Deep Reinforcement Learning for Robot Crowd Navigation21 citations · 2023
- 4
- 5
- 6A Cartesian-Based Trajectory Optimization with Jerk Constraints for a Robot10 citations · 2023
- 7
- 8
- 9
- 10