Supplementary Material: Motion Capture Recordings of Multi-Player Autonomous Racing

Motion capture recordings of autonomous quadrotor races comparing a multi-agent reinforcement learning (MARL) policy trained with league-play against a human expert pilot.本数据集为自主四旋翼无人机竞速赛事的动作捕捉（Motion Capture）录制数据，用于对比采用联赛训练法（league-play）训练得到的多智能体强化学习（multi-agent reinforcement learning, MARL）策略与人类专家飞行员的竞速表现。