Multimodal End-to-End Learning for Autonomous Steering in Adverse Road and Weather Conditions

TL;DR

This paper develops a multimodal end-to-end learning approach for autonomous steering in challenging adverse conditions, utilizing camera and lidar data to improve performance in poor visibility scenarios.

Contribution

It extends end-to-end autonomous driving models to incorporate lidar data, demonstrating improved steering prediction in adverse weather and road conditions.

Findings

Lidar data enhances model accuracy in adverse conditions

Multimodal sensor fusion outperforms single modality models

On-road tests confirm lidar's positive impact on steering prediction

Abstract

Autonomous driving is challenging in adverse road and weather conditions in which there might not be lane lines, the road might be covered in snow and the visibility might be poor. We extend the previous work on end-to-end learning for autonomous steering to operate in these adverse real-life conditions with multimodal data. We collected 28 hours of driving data in several road and weather conditions and trained convolutional neural networks to predict the car steering wheel angle from front-facing color camera images and lidar range and reflectance data. We compared the CNN model performances based on the different modalities and our results show that the lidar modality improves the performances of different multimodal sensor-fusion models. We also performed on-road tests with different models and they support this observation.