Cross-Domain Object Detection Using Unsupervised Image Translation
This addresses the problem of adapting object detectors to unseen domains for applications like autonomous driving, with incremental improvements in performance and interpretability.
The paper tackles unsupervised domain adaptation for object detection by generating an artificial dataset in the target domain using unsupervised image translation, achieving significant improvements that outperform state-of-the-art methods and close the gap toward the upper-bound in autonomous driving scenarios.
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.