Probabilistic Contrastive Learning for Domain Adaptation

TL;DR

This paper introduces Probabilistic Contrastive Learning (PCL), a novel approach that replaces features with probabilities to improve domain adaptation performance across various visual tasks.

Contribution

The paper proposes PCL, which removes $ ext{l}_2$ normalization and aligns feature probabilities with class weights, enhancing domain adaptation effectiveness.

Findings

PCL achieves consistent improvements on five domain adaptation tasks.

On SYNTHIA semantic segmentation, PCL outperforms CPSL-D by over 2% in mean IoU.

PCL requires significantly less training time compared to previous methods.

Abstract

Contrastive learning has shown impressive success in enhancing feature discriminability for various visual tasks in a self-supervised manner, but the standard contrastive paradigm (features+$\ell_{2}$ normalization) has limited benefits when applied in domain adaptation. We find that this is mainly because the class weights (weights of the final fully connected layer) are ignored in the domain adaptation optimization process, which makes it difficult for features to cluster around the corresponding class weights. To solve this problem, we propose the \emph{simple but powerful} Probabilistic Contrastive Learning (PCL), which moves beyond the standard paradigm by removing $\ell_{2}$ normalization and replacing the features with probabilities. PCL can guide the probability distribution towards a one-hot configuration, thus minimizing the discrepancy between features and class weights. We conduct extensive experiments to validate the effectiveness of PCL and observe consistent performance gains on five tasks, i.e., Unsupervised/Semi-Supervised Domain Adaptation (UDA/SSDA), Semi-Supervised Learning (SSL), UDA Detection and Semantic Segmentation. Notably, for UDA Semantic Segmentation on SYNTHIA, PCL surpasses the sophisticated CPSL-D by $>\!2\%$ in terms of mean IoU with a much lower training cost (PCL: 1*3090, 5 days v.s. CPSL-D: 4*V100, 11 days). Code is available at https://github.com/ljjcoder/Probabilistic-Contrastive-Learning.