Max Robotics — Global Service Robot Directory + US Certification

About

Mingyi He is a multidisciplinary researcher whose work spans computer vision, robotics, human motion sensing, and rehabilitation engineering. His research bridges cutting-edge machine learning techniques with real-world medical and robotic applications, making meaningful contributions across several rapidly advancing fields. In computer vision, He developed CFP-Net, a novel cross-form pyramid deep learning architecture for stereo matching that has garnered 21 citations, demonstrating his early expertise in autonomous driving and 3D scene reconstruction. His subsequent work on rolling shutter camera modeling further solidified his reputation in computational imaging. He has made particularly notable strides in human-robot interaction, pioneering bending sensor-based systems to digitize human motion for humanoid robots — a highly cited contribution (23 citations) addressing longstanding challenges in robotic dexterity. This work connects naturally to his rehabilitation robotics research, where he has developed advanced motion control strategies for lower-limb exoskeletons using reinforcement learning algorithms such as TD3, and applied iterative learning control for rehabilitation consistency. His bibliometric analysis of post-stroke upper limb dysfunction (24 citations) reflects a broader commitment to evidence-based rehabilitation research. Collectively, He's diverse yet interconnected portfolio positions him as an emerging and impactful voice at the intersection of intelligent systems and assistive healthcare technologies.