OIL: Observational Imitation Learning
This addresses the challenge of robust and scalable imitation learning for autonomous navigation, offering a novel method that improves performance over existing approaches, though it is incremental in its application to specific domains.
The paper tackles the problem of autonomous navigation by imitating teachers, proposing Observational Imitation Learning (OIL) to overcome sensitivity to teacher mistakes and poor scalability, and demonstrates that the trained network outperforms teachers, imitation learning, reinforcement learning baselines, and humans in simulation for autonomous driving and UAV racing tasks.
Recent work has explored the problem of autonomous navigation by imitating a teacher and learning an end-to-end policy, which directly predicts controls from raw images. However, these approaches tend to be sensitive to mistakes by the teacher and do not scale well to other environments or vehicles. To this end, we propose Observational Imitation Learning (OIL), a novel imitation learning variant that supports online training and automatic selection of optimal behavior by observing multiple imperfect teachers. We apply our proposed methodology to the challenging problems of autonomous driving and UAV racing. For both tasks, we utilize the Sim4CV simulator that enables the generation of large amounts of synthetic training data and also allows for online learning and evaluation. We train a perception network to predict waypoints from raw image data and use OIL to train another network to predict controls from these waypoints. Extensive experiments demonstrate that our trained network outperforms its teachers, conventional imitation learning (IL) and reinforcement learning (RL) baselines and even humans in simulation. The project website is available at https://sites.google.com/kaust.edu.sa/oil/ and a video at https://youtu.be/_rhq8a0qgeg