Approximate Inverse Reinforcement Learning from Vision-based Imitation Learning

In this work, we present a method for obtaining an implicit objective\nfunction for vision-based navigation. The proposed methodology relies on\nImitation Learning, Model Predictive Control (MPC), and an interpretation\ntechnique used in Deep Neural Networks. We use Imitation Learning as a means to\ndo Inverse Reinforcement Learning in order to create an approximate cost\nfunction generator for a visual navigation challenge. The resulting cost\nfunction, the costmap, is used in conjunction with MPC for real-time control\nand outperforms other state-of-the-art costmap generators in novel\nenvironments. The proposed process allows for simple training and robustness to\nout-of-sample data. We apply our method to the task of vision-based autonomous\ndriving in multiple real and simulated environments and show its\ngeneralizability.\n