Latent Space Reinforcement Learning for Steering Angle Prediction
This addresses the problem of autonomous navigation for self-driving cars, but it is incremental as it builds upon existing deep reinforcement learning approaches.
The paper tackled the problem of learning driving policies for an autonomous agent in a high-fidelity simulator by predicting steering angles from raw images, and the result showed that the method could learn to maneuver the car without human control signals.
Model-free reinforcement learning has recently been shown to successfully learn navigation policies from raw sensor data. In this work, we address the problem of learning driving policies for an autonomous agent in a high-fidelity simulator. Building upon recent research that applies deep reinforcement learning to navigation problems, we present a modular deep reinforcement learning approach to predict the steering angle of the car from raw images. The first module extracts a low-dimensional latent semantic representation of the image. The control module trained with reinforcement learning takes the latent vector as input to predict the correct steering angle. The experimental results have showed that our method is capable of learning to maneuver the car without any human control signals.