Assessing Domain Gap for Continual Domain Adaptation in Object Detection

TL;DR

This paper proposes a selective domain adaptation method for object detection that uses domain gap metrics to decide when to adapt, improving efficiency without sacrificing accuracy in changing environments.

Contribution

It introduces a domain gap-based criterion for selective adaptation in continual object detection, reducing computational costs while maintaining performance.

Findings

Domain gap correlates with detection accuracy.

Selective adaptation improves efficiency in cyclical environmental changes.

Method maintains detection performance with less frequent updates.

Abstract

To ensure reliable object detection in autonomous systems, the detector must be able to adapt to changes in appearance caused by environmental factors such as time of day, weather, and seasons. Continually adapting the detector to incorporate these changes is a promising solution, but it can be computationally costly. Our proposed approach is to selectively adapt the detector only when necessary, using new data that does not have the same distribution as the current training data. To this end, we investigate three popular metrics for domain gap evaluation and find that there is a correlation between the domain gap and detection accuracy. Therefore, we apply the domain gap as a criterion to decide when to adapt the detector. Our experiments show that our approach has the potential to improve the efficiency of the detector's operation in real-world scenarios, where environmental conditions change in a cyclical manner, without sacrificing the overall performance of the detector. Our code is publicly available at https://github.com/dadung/DGE-CDA.