SoloParkour: 特権経験を用いた視覚移動のための制約付き強化学習
SoloParkour: Constrained Reinforcement Learning for Visual Locomotion from Privileged Experience
特権情報で訓練した方策から経験を生成し、深度画像からのオフポリシー強化学習を効率化することで、四足ロボットに歩行・登攀・跳躍・這行などのパルクール技能を実現した。
著者: Elliot Chane-Sane, Joseph Amigo, Thomas Flayols, Ludovic Righetti, Nicolas Mansard
分類: cs.RO, cs.CV, cs.LG
原文アブストラクト
Parkour poses a significant challenge for legged robots, requiring navigation through complex environments with agility and precision based on limited sensory inputs. In this work, we introduce a novel method for training end-to-end visual policies, from depth pixels to robot control commands, to achieve agile and safe quadruped locomotion. We formulate robot parkour as a constrained reinforcement learning (RL) problem designed to maximize the emergence of agile skills within the robot's physical limits while ensuring safety. We first train a policy without vision using privileged information about the robot's surroundings. We then generate experience from this privileged policy to warm-start a sample efficient off-policy RL algorithm from depth images. This allows the robot to adapt behaviors from this privileged experience to visual locomotion while circumventing the high computational costs of RL directly from pixels. We demonstrate the effectiveness of our method on a real Solo-12 robot, showcasing its capability to perform a variety of parkour skills such as walking, climbing, leaping, and crawling.
