Improving Cross-Domain Detection with Self-Supervised Learning

• Published in: 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) pp. 4746 - 4755
• Main Authors: Li, Kai, Wigington, Curtis, Tensmeyer, Chris, Morariu, Vlad I., Zhao, Handong, Manjunatha, Varun, Barmpalios, Nikolaos, Fu, Yun
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.06.2023
• Subjects: Adaptation models; Data augmentation; Data models; Detectors; Pattern recognition; Predictive models; Self-supervised learning
• ISSN: 2160-7516
• DOI: 10.1109/CVPRW59228.2023.00503

Cross-Domain Detection (XDD) aims to train a domain-adaptive object detector using unlabeled images from a target domain and labeled images from a source domain. Existing approaches achieve this either by transferring the style of source images to that of target images, or by aligning the features of images from the two domains. In this paper, rather than proposing another method following the existing lines, we introduce a new framework complementary to existing methods. Our framework unifies some popular Self-Supervised Learning (SSL) techniques (e.g., rotation angle prediction, strong/weak data augmentation, mean teacher modeling) and adapts them to the XDD task. Our basic idea is to leverage the unsupervised nature of these SSL techniques and apply them simultaneously across domains (source and target) and models (student and teacher). These SSL techniques can thus serve as shared bridges that facilitate knowledge transfer between domains. More importantly, as these techniques are independently applied in each domain, they are complementary to existing domain alignment techniques that relies on interactions between domains (e.g., adversarial alignment). We perform extensive analyses on these SSL techniques and show that they significantly improve the performance of existing methods. In addition, we reach comparable or even better performance than the state-of-the-art methods when integrating our framework with an old well-established method.