Unsupervised domain adaptation by learning using privileged information

TL;DR

This paper introduces a novel approach to unsupervised domain adaptation that leverages privileged information available only during training to improve transfer accuracy, robustness, and sample efficiency in high-dimensional tasks like image classification.

Contribution

It proposes a two-stage learning algorithm utilizing privileged information to relax input restrictions and enhance domain adaptation performance, with practical end-to-end implementation for images.

Findings

Privileged information reduces domain transfer errors.

Method is robust to spurious correlations.

Sample efficiency is significantly improved.

Abstract

Successful unsupervised domain adaptation is guaranteed only under strong assumptions such as covariate shift and overlap between input domains. The latter is often violated in high-dimensional applications like image classification which, despite this limitation, continues to serve as inspiration and benchmark for algorithm development. In this work, we show that training-time access to side information in the form of auxiliary variables can help relax restrictions on input variables and increase the sample efficiency of learning at the cost of collecting a richer variable set. As this information is assumed available only during training, not in deployment, we call this problem unsupervised domain adaptation by learning using privileged information (DALUPI). To solve this problem, we propose a simple two-stage learning algorithm, inspired by our analysis of the expected error in the target domain, and a practical end-to-end variant for image classification. We propose three evaluation tasks based on classification of entities in photos and anomalies in medical images with different types of available privileged information (binary attributes and single or multiple regions of interest). We demonstrate across these tasks that using privileged information in learning can reduce errors in domain transfer compared to baselines, be robust to spurious correlations in the source domain, and increase sample efficiency.