Reducing Domain Gap by Reducing Style Bias

• Published in: 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) pp. 8686 - 8695
• Main Authors: Nam, Hyeonseob, Lee, HyunJae, Park, Jongchan, Yoon, Wonjun, Yoo, Donggeun
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.06.2021
• Subjects: Adaptation models; Aerospace electronics; Computer vision; Decision making; Encoding; Robustness; Shape

Convolutional Neural Networks (CNNs) often fail to maintain their performance when they confront new test domains, which is known as the problem of domain shift. Recent studies suggest that one of the main causes of this problem is CNNs' strong inductive bias towards image styles (i.e. textures...