A Review of Single-Source Deep Unsupervised Visual Domain Adaptation

TL;DR

This paper reviews recent single-source deep unsupervised visual domain adaptation methods, highlighting their strategies, datasets, and future research challenges to improve model transfer across different visual domains.

Contribution

It provides a comprehensive overview and comparison of current methods in single-source unsupervised domain adaptation for visual tasks, and discusses future research directions.

Findings

Summarizes discrepancy-based, adversarial, and self-supervision methods.

Analyzes benchmark datasets used in domain adaptation.

Identifies challenges and potential solutions for future research.

Abstract

Large-scale labeled training datasets have enabled deep neural networks to excel across a wide range of benchmark vision tasks. However, in many applications, it is prohibitively expensive and time-consuming to obtain large quantities of labeled data. To cope with limited labeled training data, many have attempted to directly apply models trained on a large-scale labeled source domain to another sparsely labeled or unlabeled target domain. Unfortunately, direct transfer across domains often performs poorly due to the presence of domain shift or dataset bias. Domain adaptation is a machine learning paradigm that aims to learn a model from a source domain that can perform well on a different (but related) target domain. In this paper, we review the latest single-source deep unsupervised domain adaptation methods focused on visual tasks and discuss new perspectives for future research. We begin with the definitions of different domain adaptation strategies and the descriptions of existing benchmark datasets. We then summarize and compare different categories of single-source unsupervised domain adaptation methods, including discrepancy-based methods, adversarial discriminative methods, adversarial generative methods, and self-supervision-based methods. Finally, we discuss future research directions with challenges and possible solutions.