Spatial Class Distribution Shift in Unsupervised Domain Adaptation: Local Alignment Comes to Rescue

Safa Cicek, Ning Xu, Zhaowen Wang, Hailin Jin, Stefano Soatto; Proceedings of the Asian Conference on Computer Vision (ACCV), 2020

Abstract

We propose a method for semantic segmentation in unsupervised domain adaptation (UDA) setting. We particularly examine the domain gap between spatial-class distributions and propose to align the local distributions of the segmentation predictions. Despite its simplicity, the proposed method achieves state-of-the-art results in UDA segmentation benchmarks.

Related Material

[pdf] [supp]
[bibtex]
@InProceedings{Cicek_2020_ACCV, author = {Cicek, Safa and Xu, Ning and Wang, Zhaowen and Jin, Hailin and Soatto, Stefano}, title = {Spatial Class Distribution Shift in Unsupervised Domain Adaptation: Local Alignment Comes to Rescue}, booktitle = {Proceedings of the Asian Conference on Computer Vision (ACCV)}, month = {November}, year = {2020} }