Towards Closing the Domain Gap with Event Cameras

TL;DR

This paper investigates using event cameras to address the domain gap caused by lighting differences in end-to-end driving, demonstrating their robustness across day-night conditions compared to traditional cameras.

Contribution

The study proposes event cameras as an alternative to traditional cameras for reducing domain gap issues, showing their effectiveness in maintaining performance across lighting conditions.

Findings

Event cameras exhibit smaller domain-shift penalties than grayscale frames.

Event cameras provide more consistent performance across day-night lighting changes.

Baseline performance with event cameras surpasses traditional camera-based methods in cross-domain scenarios.

Abstract

Although traditional cameras are the primary sensor for end-to-end driving, their performance suffers greatly when the conditions of the data they were trained on does not match the deployment environment, a problem known as the domain gap. In this work, we consider the day-night lighting difference domain gap. Instead of traditional cameras we propose event cameras as a potential alternative which can maintain performance across lighting condition domain gaps without requiring additional adjustments. Our results show that event cameras maintain more consistent performance across lighting conditions, exhibiting domain-shift penalties that are generally comparable to or smaller than grayscale frames and provide superior baseline performance in cross-domain scenarios.