Domain-Specific Suppression for Adaptive Object Detection

Yu Wang, Rui Zhang, Shuo Zhang, Miao Li, Yangyang Xia, Xishan Zhang, Shaoli Liu; Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2021, pp. 9603-9612


Domain adaptation methods face performance degradation in object detection, as the complexity of tasks require more about the transferability of the model. We propose a new perspective on how CNN models gain the transferability, viewing the weights of a model as a series of motion patterns. The directions of weights, and the gradients, can be divided into domain-specific and domain-invariant parts, and the goal of domain adaptation is to concentrate on the domain-invariant direction while eliminating the disturbance from domain-specific one. Current UDA object detection methods view the two directions as a whole while optimizing, which will cause domain-invariant direction mismatch even if the output features are perfectly aligned. In this paper, we propose the domain-specific suppression, an exemplary and generalizable constraint to the original convolution gradients in backpropagation to detach the two parts of directions and suppress the domain-specific one. We further validate our theoretical analysis and methods on several domain adaptive object detection tasks, including weather, camera configuration, and synthetic to real-world adaptation. Our experiment results show significant advance over the state-of-the-art methods in the UDA object detection field, performing a promotion of 10.2 12.2% mAP on all these domain adaptation scenarios.

Related Material

[pdf] [arXiv]
@InProceedings{Wang_2021_CVPR, author = {Wang, Yu and Zhang, Rui and Zhang, Shuo and Li, Miao and Xia, Yangyang and Zhang, Xishan and Liu, Shaoli}, title = {Domain-Specific Suppression for Adaptive Object Detection}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month = {June}, year = {2021}, pages = {9603-9612} }