End-to-end learning for lane keeping of self-driving cars

TLDR

Lane keeping is a critical feature for autonomous vehicles, and unlike traditional modular approaches, an end‑to‑end model can steer directly from front‑view camera data after training. The study proposes an end‑to‑end learning approach to determine the correct steering angle for lane keeping. A convolutional neural network processes raw front‑view images to output steering angles, trained and evaluated on the comma.ai dataset of driving footage and corresponding steering data. The model learns to maintain lane position from human driving data, and the paper discusses its performance and limitations.

Abstract

Lane keeping is an important feature for self-driving cars. This paper presents an end-to-end learning approach to obtain the proper steering angle to maintain the car in the lane. The convolutional neural network (CNN) model takes raw image frames as input and outputs the steering angles accordingly. The model is trained and evaluated using the comma.ai dataset, which contains the front view image frames and the steering angle data captured when driving on the road. Unlike the traditional approach that manually decomposes the autonomous driving problem into technical components such as lane detection, path planning and steering control, the end-to-end model can directly steer the vehicle from the front view camera data after training. It learns how to keep in lane from human driving data. Further discussion of this end-to-end approach and its limitation are also provided.