Deep Learning and Control Algorithms of Direct Perception for Autonomous Driving

TL;DR

This paper enhances autonomous driving perception by modifying CNNs to estimate key driving indicators and designing a controller that uses these indicators for collision avoidance in simulation.

Contribution

It introduces modified CNN architectures for direct perception and a control algorithm using these perceptions in a simulated environment, improving efficiency and stability.

Findings

CNN modifications outperform earlier algorithms in training efficiency

The perception-controller system achieves better driving stability

The approach is validated in diverse simulated traffic scenarios

Abstract

Based on the direct perception paradigm of autonomous driving, we investigate and modify the CNNs (convolutional neural networks) AlexNet and GoogLeNet that map an input image to few perception indicators (heading angle, distances to preceding cars, and distance to road centerline) for estimating driving affordances in highway traffic. We also design a controller with these indicators and the short-range sensor information of TORCS (the open racing car simulator) for driving simulated cars to avoid collisions. We collect a set of images from a TORCS camera in various driving scenarios, train these CNNs using the dataset, test them in unseen traffics, and find that they perform better than earlier algorithms and controllers in terms of training efficiency and driving stability. Source code and data are available on our website.