Max Robotics — Global Service Robot Directory + US Certification

About

Hongzhuo Liang is a robotics researcher whose work spans dexterous robotic manipulation, teleoperation, reinforcement learning, and multimodal sensing. He is perhaps best known for pioneering vision-based teleoperation systems for dexterous robotic hands, introducing TeachNet (2019, 104 citations), an end-to-end deep neural network that translates depth images of human hands into robot joint configurations without markers or wearable devices. This foundational work was extended through subsequent systems incorporating IMU-based arm tracking and active vision to address finger occlusion challenges, collectively garnering over 130 additional citations. Liang's research also makes significant contributions to multimodal robot learning, combining tactile sensing, proprioception, and vision to enable anthropomorphic hands to grasp novel objects (2021, 45 citations) and to bridge the sim-to-real gap in contact-rich assembly tasks (2023, 47 citations). His reinforcement learning work explores pushing-then-grasping strategies, planar pushing with vision-proprioception models, and task-oriented 6-DoF grasp detection, reflecting a broad commitment to practical, intelligent robot autonomy. With over 400 cumulative citations across diverse manipulation domains, Liang has established himself as a productive contributor to the intersection of deep learning and embodied robotics, offering tools and insights directly applicable to smart manufacturing and human-robot collaboration.