CLUDA : Contrastive Learning in Unsupervised Domain Adaptation for Semantic Segmentation

2022-08-27 · Midhun Vayyat, Jaswin Kasi, Anuraag Bhattacharya, Shuaib Ahmed, Rahul Tallamraju

In this work, we propose CLUDA, a simple, yet novel method for performing unsupervised domain adaptation (UDA) for semantic segmentation by incorporating contrastive losses into a student-teacher learning paradigm, that makes use of pseudo-labels generated from the target domain by the teacher network. More specifically, we extract a multi-level fused-feature map from the encoder, and apply contrastive loss across different classes and different domains, via source-target mixing of images. We consistently improve performance on various feature encoder architectures and for different domain adaptation datasets in semantic segmentation. Furthermore, we introduce a learned-weighted contrastive loss to improve upon on a state-of-the-art multi-resolution training approach in UDA. We produce state-of-the-art results on GTA $\rightarrow$ Cityscapes (74.4 mIOU, +0.6) and Synthia $\rightarrow$ Cityscapes (67.2 mIOU, +1.4) datasets. CLUDA effectively demonstrates contrastive learning in UDA as a generic method, which can be easily integrated into any existing UDA for semantic segmentation tasks. Please refer to the supplementary material for the details on implementation.

📄 PDF Abstract BibTeX arXiv:2208.14227

Code (1)

user0407/CLUDA 공식 구현 pytorch

Tasks

Contrastive Learning Domain Adaptation Segmentation Semantic Segmentation Synthetic-to-Real Translation Unsupervised Domain Adaptation

Methods 이 논문이 사용한 방법론

Contrastive Learning 설명 없음

Contrastive Learning for Unsupervised Domain Adaptation of Time Series

2022-06-13 · Yilmazcan Ozyurt, Stefan Feuerriegel, Ce Zhang

Unsupervised domain adaptation (UDA) aims at learning a machine learning model using a labeled source domain that performs well on a similar yet different, unlabeled target domain. UDA is important in many applications s…

Contrastive Learning Domain Adaptation Time Series Time Series Analysis+1

Lifelong Unsupervised Domain Adaptive Person Re-identification with Coordinated Anti-forgetting and Adaptation

2021-12-13 · CVPR 2022 1 · Zhipeng Huang, Zhizheng Zhang, Cuiling Lan, Wenjun Zeng 외

Unsupervised domain adaptive person re-identification (ReID) has been extensively investigated to mitigate the adverse effects of domain gaps. Those works assume the target domain data can be accessible all at once. Howe…

Domain Adaptive Person Re-Identification Knowledge Distillation Memorization Person Re-Identification+2

Unsupervised Contrastive Domain Adaptation for Semantic Segmentation

2022-04-18 · Feihu Zhang, Vladlen Koltun, Philip Torr, René Ranftl 외

Semantic segmentation models struggle to generalize in the presence of domain shift. In this paper, we introduce contrastive learning for feature alignment in cross-domain adaptation. We assemble both in-domain contrasti…

Contrastive Learning Domain Adaptation Semantic Segmentation

Domain Adaptation for Semantic Segmentation via Patch-Wise Contrastive Learning

2021-04-22 · Weizhe Liu, David Ferstl, Samuel Schulter, Lukas Zebedin 외

We introduce a novel approach to unsupervised and semi-supervised domain adaptation for semantic segmentation. Unlike many earlier methods that rely on adversarial learning for feature alignment, we leverage contrastive …

Contrastive Learning Domain Adaptation Segmentation Semantic Segmentation+1

Contrast and Mix: Temporal Contrastive Video Domain Adaptation with Background Mixing

2021-10-28 · NeurIPS 2021 12 · Aadarsh Sahoo, Rutav Shah, Rameswar Panda, Kate Saenko 외

Unsupervised domain adaptation which aims to adapt models trained on a labeled source domain to a completely unlabeled target domain has attracted much attention in recent years. While many domain adaptation techniques h…

Contrastive Learning Domain Adaptation Unsupervised Domain Adaptation Video Domain Adapation