Domain Adaptation for Visual Applications: A Comprehensive Survey

TL;DR

This comprehensive survey reviews the evolution of domain adaptation techniques in visual applications, covering shallow and deep methods, various scenarios, and their relation to broader transfer learning approaches.

Contribution

It provides an extensive overview of both traditional and deep domain adaptation methods across multiple visual tasks, highlighting recent advances and future directions.

Findings

Deep architectures have significantly advanced domain adaptation effectiveness.

Various methods address both homogeneous and heterogeneous domain shifts.

Domain adaptation techniques are now applied beyond image classification to complex visual tasks.

Abstract

The aim of this paper is to give an overview of domain adaptation and transfer learning with a specific view on visual applications. After a general motivation, we first position domain adaptation in the larger transfer learning problem. Second, we try to address and analyze briefly the state-of-the-art methods for different types of scenarios, first describing the historical shallow methods, addressing both the homogeneous and the heterogeneous domain adaptation methods. Third, we discuss the effect of the success of deep convolutional architectures which led to new type of domain adaptation methods that integrate the adaptation within the deep architecture. Fourth, we overview the methods that go beyond image categorization, such as object detection or image segmentation, video analyses or learning visual attributes. Finally, we conclude the paper with a section where we relate domain adaptation to other machine learning solutions.