Learning Unbiased Rewards with Mutual Information in Adversarial Imitation Learning | IEEE Conference Publication | IEEE Xplore