Con$^{2}$DA: Simplifying Semi-supervised Domain Adaptation by Learning Consistent and Contrastive Feature Representations

TL;DR

Con$^{2}$DA introduces a simple yet effective semi-supervised domain adaptation framework that learns consistent and contrastive feature representations through data augmentation, achieving state-of-the-art results on benchmark datasets.

Contribution

The paper proposes Con$^{2}$DA, a novel framework that extends semi-supervised learning techniques to domain adaptation by enforcing consistency and contrastiveness in feature representations.

Findings

Learning consistent and contrastive features improves domain adaptation.

Strong data augmentation enhances model performance.

Achieves state-of-the-art results on benchmark datasets.

Abstract

In this work, we present Con$^{2}$DA, a simple framework that extends recent advances in semi-supervised learning to the semi-supervised domain adaptation (SSDA) problem. Our framework generates pairs of associated samples by performing stochastic data transformations to a given input. Associated data pairs are mapped to a feature representation space using a feature extractor. We use different loss functions to enforce consistency between the feature representations of associated data pairs of samples. We show that these learned representations are useful to deal with differences in data distributions in the domain adaptation problem. We performed experiments to study the main components of our model and we show that (i) learning of the consistent and contrastive feature representations is crucial to extract good discriminative features across different domains, and ii) our model benefits from the use of strong augmentation policies. With these findings, our method achieves state-of-the-art performances in three benchmark datasets for SSDA.