FastMimic: Model-Based Motion Imitation for Agile, Diverse and Generalizable Quadrupedal Locomotion

Robots operating in human environments require a diverse set of skills, including slow and fast walking, turning, side-stepping, more. However, developing robot controllers capable exhibiting such broad range behaviors is challenging problem that necessitates meticulous investigation for each task. To address this challenge, we introduce trajectory optimization method resolves the kinematic infeasibility reference animal motions. This method, combined with model-based controller, results unified data-driven control framework imitating various gaits without need expensive simulation training or real-world fine-tuning. Our variety motor skills as trotting, pacing, side-stepping ease. It shows superior tracking capabilities both simulations real world compared to other imitation controllers, one learning-based motion technique.