Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Cross-domain Contrastive Learning for Unsupervised Domain Adaptation

About

Unsupervised domain adaptation (UDA) aims to transfer knowledge learned from a fully-labeled source domain to a different unlabeled target domain. Most existing UDA methods learn domain-invariant feature representations by minimizing feature distances across domains. In this work, we build upon contrastive self-supervised learning to align features so as to reduce the domain discrepancy between training and testing sets. Exploring the same set of categories shared by both domains, we introduce a simple yet effective framework CDCL, for domain alignment. In particular, given an anchor image from one domain, we minimize its distances to cross-domain samples from the same class relative to those from different categories. Since target labels are unavailable, we use a clustering-based approach with carefully initialized centers to produce pseudo labels. In addition, we demonstrate that CDCL is a general framework and can be adapted to the data-free setting, where the source data are unavailable during training, with minimal modification. We conduct experiments on two widely used domain adaptation benchmarks, i.e., Office-31 and VisDA-2017, for image classification tasks, and demonstrate that CDCL achieves state-of-the-art performance on both datasets.

Rui Wang, Zuxuan Wu, Zejia Weng, Jingjing Chen, Guo-Jun Qi, Yu-Gang Jiang• 2021

Related benchmarks

TaskDatasetResultRank
ClassificationC16
Balanced Accuracy66.6
31
ClassificationCAMELYON17 Center 3
CL Score77.4
18
ClassificationCAMELYON17 Center 0
Classification Accuracy72.5
18
ClassificationCAMELYON17 Center 1
CL59.3
18
ClassificationCAMELYON17 Center 2
CL72.8
18
ClassificationCAMELYON17 Center 4
CL Score55.4
18
LocalizationC16 Target Domain from GlaS → C16
PxAP37.9
18
LocalizationCAMELYON17 Center 0
PxAP27.2
18
LocalizationCAMELYON17 Center 3
PxAP37.6
18
ClassificationAverage C16, C17-0, C17-1, C17-2, C17-3, C17-4
Accuracy (CL Average)66.9
18
Showing 10 of 14 rows

Other info

Follow for update