Contrast, Stylize and Adapt: Unsupervised Contrastive Learning Framework for Domain Adaptive Semantic Segmentation

Tianyu Li, Subhankar Roy, Huayi Zhou, Hongtao Lu, Stéphane Lathuilière

CVPRW 2023 pp. 4869-4879

doi:10.1109/CVPRW59228.2023.00515 /cvprw/2023/li2023cvprw-contrast/

Abstract

To overcome the domain gap between synthetic and real-world datasets, unsupervised domain adaptation methods have been proposed for semantic segmentation. Majority of the previous approaches have attempted to reduce the gap either at the pixel or feature level, disregarding the fact that the two components interact positively. To address this, we present CONtrastive FEaTure and pIxel alignment (CON-FETI) for bridging the domain gap at both the pixel and feature levels using a unique contrastive formulation. We introduce well-estimated prototypes by including category-wise cross-domain information to link the two alignments: the pixel-level alignment is achieved using the jointly trained style transfer module with the prototypical semantic consistency, while the feature-level alignment is enforced to cross-domain features with the pixel-to-prototype contrast. Our extensive experiments demonstrate that our method outperforms existing state-of-the-art methods using DeepLabV2. Our code1 has been made publicly available.

PDF CVPRW Semantic Scholar

Cite

Text

Li et al. "Contrast, Stylize and Adapt: Unsupervised Contrastive Learning Framework for Domain Adaptive Semantic Segmentation." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2023. doi:10.1109/CVPRW59228.2023.00515

Markdown

[Li et al. "Contrast, Stylize and Adapt: Unsupervised Contrastive Learning Framework for Domain Adaptive Semantic Segmentation." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2023.](https://mlanthology.org/cvprw/2023/li2023cvprw-contrast/) doi:10.1109/CVPRW59228.2023.00515

BibTeX

@inproceedings{li2023cvprw-contrast,
  title     = {{Contrast, Stylize and Adapt: Unsupervised Contrastive Learning Framework for Domain Adaptive Semantic Segmentation}},
  author    = {Li, Tianyu and Roy, Subhankar and Zhou, Huayi and Lu, Hongtao and Lathuilière, Stéphane},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
  year      = {2023},
  pages     = {4869-4879},
  doi       = {10.1109/CVPRW59228.2023.00515},
  url       = {https://mlanthology.org/cvprw/2023/li2023cvprw-contrast/}
}