This project aims to efficiently estimate complete human-like variable impedance skills from human demonstrations and to transfer the skills to real-world robots. In this project, we applied GMM-GMR to extract a reference pose trajectory and estimate stiffness profiles from collected data. Then, we utilized DMP to imitate and generalize the taught skill.
Python
This project aims to efficiently estimate complete human-like variable impedance skills from human demonstrations and to transfer the skills to real-world robots.
In this project, we applied GMM-GMR to extract a reference pose trajectory and estimate stiffness profiles from collected data. Then, we utilized DMP to imitate and generalize the taught skill.