Muhan Hou

Papers

Total Citations

H-Index

About

Muhan Hou is a robotics researcher focused on advancing human-robot interaction through intelligent learning systems. Their primary research areas include Learning from Demonstrations (LfD), active robot learning, and imitation learning for human-centered tasks. Hou’s major contributions address critical challenges in robot skill acquisition, particularly the problem of data imbalance in demonstrations. In their 2023 work "Shaping Imbalance into Balance" (4 citations), they pioneered an active robot guidance approach that helps human teachers provide more balanced demonstrations during the teaching process, rather than correcting imbalances after data collection. This proactive method represents a significant shift from traditional post-hoc correction techniques. Their 2023 framework for process-oriented imitation learning (3 citations) specifically targets human-centered interactive tasks like social greetings and cooperative dressing, emphasizing the real-time spatial and temporal coordination required between humans and robots. Most recently, their 2025 work on active robot curriculum learning (3 citations) addresses the challenge of sub-optimal teaching from untrained demonstrators by enabling robots to intelligently request demonstrations that optimize their learning trajectory. Hou’s research is particularly notable for its focus on making robot learning more robust and practical for real-world human-robot collaboration scenarios.