Zhihao Cheng

Xidian University, The University of Sydney

Papers

Total Citations

H-Index

About

Zhihao Cheng is a robotics researcher whose work bridges the gap between human motion and robotic imitation, with a focus on enabling humanoid robots to learn and replicate complex behaviors. His key research areas include imitation learning, human-robot interaction, and whole-body motion control. Cheng’s major contribution is the development of a real-time algorithm that allows humanoid robots to mimic eight-chain whole-body motions from human skeletal data captured by a Kinect sensor—a first in the field, as detailed in his 2016 paper (22 citations). This work laid the foundation for more natural and intuitive robot control. He has also advanced the theoretical understanding of imitation learning, notably demonstrating the guaranteed almost equivalence between learning from observation (LfO) and learning from demonstration (LfD) in his 2021 paper (8 citations), a finding that simplifies robot training by eliminating the need for expert action data. With a growing citation impact, Cheng’s research is shaping the future of autonomous, human-like robotics, making him a notable figure in the pursuit of seamless human-robot collaboration.