记忆状态评论家：面向视觉追逃的非对称Actor-Critic
原标题：Memory-State Critic for Asymmetric Actor-Critic with Application to Vision-Based Pursuit-Evasion
AI 摘要
该论文提出一种名为 memory-state critic 的非对称 actor-critic 方法，让 critic 以环境状态和策略自身的循环记忆为条件，而无需第二个历史编码器。作者证明该方法与 history-state critic 具有相同的 on-policy 价值函数和无偏策略梯度，且无需将 critic 损失反传进策略记忆。在双四旋翼视觉追逃环境中，该方法比 history-state critic 收敛更快，并在 wall arena 中略优于仅用状态的 critic。
正文节选
Memory-State Critic for Asymmetric Actor-Critic with Application to Vision-Based Pursuit-Evasion Abstract In partially observable Markov decision processes, the optimal policy generally depends on the history of observations and past actions. Asymmetric actor-critic methods have become popular to learn such policies when additional information, such as the true state of the environment, is available during training. The critic, which is not needed at execution, is given access to the state. A cr