Cross-Domain Object Detection Using Unsupervised Image Translation

TL;DR

This paper introduces a less complex, more interpretable method for unsupervised domain adaptation in object detection by generating artificial target domain datasets using image translation, improving performance in autonomous driving scenarios.

Contribution

Proposes a novel approach using unsupervised image translation to generate artificial target domain data, enhancing object detection without complex feature alignment methods.

Findings

Outperforms state-of-the-art methods in autonomous driving scenarios

Significantly improves detection accuracy in target domain

Closes the performance gap towards the upper-bound

Abstract

Unsupervised domain adaptation for object detection addresses the adaption of detectors trained in a source domain to work accurately in an unseen target domain. Recently, methods approaching the alignment of the intermediate features proven to be promising, achieving state-of-the-art results. However, these methods are laborious to implement and hard to interpret. Although promising, there is still room for improvements to close the performance gap toward the upper-bound (when training with the target data). In this work, we propose a method to generate an artificial dataset in the target domain to train an object detector. We employed two unsupervised image translators (CycleGAN and an AdaIN-based model) using only annotated data from the source domain and non-annotated data from the target domain. Our key contributions are the proposal of a less complex yet more effective method that also has an improved interpretability. Results on real-world scenarios for autonomous driving show significant improvements, outperforming state-of-the-art methods in most cases, further closing the gap toward the upper-bound.