Active Perception in Adversarial Scenarios using Maximum Entropy Deep Reinforcement Learning | IEEE Conference Publication | IEEE Xplore