SCL: Towards Accurate Domain Adaptive Object Detection via Gradient Detach Based Stacked Complementary LossesDownload PDF

25 Sep 2019 (modified: 24 Dec 2019)ICLR 2020 Conference Withdrawn SubmissionReaders: Everyone
  • Original Pdf: pdf
  • TL;DR: We introduce a new gradient detach based complementary objective training strategy for domain adaptive object detection.
  • Abstract: Unsupervised domain adaptive object detection aims to learn a robust detector on the domain shift circumstance, where the training (source) domain is label-rich with bounding box annotations, while the testing (target) domain is label-agnostic and the feature distributions between training and testing domains are dissimilar or even totally different. In this paper, we propose a gradient detach based Stacked Complementary Losses (SCL) method that uses detection objective (cross entropy and smooth l1 regression) as the primary objective, and cuts in several auxiliary losses in different network stages to utilize information from the complement data (target images) that can be effective in adapting model parameters to both source and target domains. A gradient detach operation is applied between detection and context sub-networks during training to force networks to learn discriminative representations. We argue that the conventional training with primary objective mainly leverages the information from the source-domain for maximizing likelihood and ignores the complement data in shallow layers of networks, which leads to an insufficient integration within different domains. Thus, our proposed method is a more syncretic adaptation learning process. We conduct comprehensive experiments on seven datasets, the results demonstrate that our method performs favorably better than the state-of-the-art methods by a large margin. For instance, from Cityscapes to FoggyCityscapes, we achieve 37.9% mAP, outperforming the previous art Strong-Weak by 3.6%.
  • Keywords: Domain Adaptation, Object Detection, Gradient Detach, Stacked Complementary Losses
  • Code:
10 Replies