Key Design Choices for Double-Transfer in Source-Free Unsupervised Domain Adaptation

TL;DR

This paper thoroughly analyzes key design choices in Source-Free Unsupervised Domain Adaptation (SF-UDA), identifying critical factors and proposing effective strategies based on extensive empirical evaluation across numerous models and domain pairs.

Contribution

It provides the first comprehensive empirical study of SF-UDA, highlighting the importance of normalization, pre-training, and architecture, and offers practical recipes for improved performance.

Findings

Normalization approach is critical for SF-UDA.

Pre-training strategy significantly impacts adaptation success.

SF-UDA can match UDA performance with less data and computation.

Abstract

Fine-tuning and Domain Adaptation emerged as effective strategies for efficiently transferring deep learning models to new target tasks. However, target domain labels are not accessible in many real-world scenarios. This led to the development of Unsupervised Domain Adaptation (UDA) methods, which only employ unlabeled target samples. Furthermore, efficiency and privacy requirements may also prevent the use of source domain data during the adaptation stage. This challenging setting, known as Source-Free Unsupervised Domain Adaptation (SF-UDA), is gaining interest among researchers and practitioners due to its potential for real-world applications. In this paper, we provide the first in-depth analysis of the main design choices in SF-UDA through a large-scale empirical study across 500 models and 74 domain pairs. We pinpoint the normalization approach, pre-training strategy, and backbone architecture as the most critical factors. Based on our quantitative findings, we propose recipes to best tackle SF-UDA scenarios. Moreover, we show that SF-UDA is competitive also beyond standard benchmarks and backbone architectures, performing on par with UDA at a fraction of the data and computational cost. In the interest of reproducibility, we include the full experimental results and code as supplementary material.