Domain Generalization by Learning and Removing Domain-specific Features

TL;DR

This paper introduces LRDG, a framework that explicitly removes domain-specific features from images to improve the generalization ability of deep neural networks across unseen domains.

Contribution

The paper proposes a novel encoder-decoder based framework that learns to remove domain-specific features, enhancing domain generalization in deep neural networks.

Findings

LRDG outperforms state-of-the-art methods on benchmark datasets.

Explicit removal of domain-specific features improves model robustness.

The approach effectively learns domain-invariant features for better generalization.

Abstract

Deep Neural Networks (DNNs) suffer from domain shift when the test dataset follows a distribution different from the training dataset. Domain generalization aims to tackle this issue by learning a model that can generalize to unseen domains. In this paper, we propose a new approach that aims to explicitly remove domain-specific features for domain generalization. Following this approach, we propose a novel framework called Learning and Removing Domain-specific features for Generalization (LRDG) that learns a domain-invariant model by tactically removing domain-specific features from the input images. Specifically, we design a classifier to effectively learn the domain-specific features for each source domain, respectively. We then develop an encoder-decoder network to map each input image into a new image space where the learned domain-specific features are removed. With the images output by the encoder-decoder network, another classifier is designed to learn the domain-invariant features to conduct image classification. Extensive experiments demonstrate that our framework achieves superior performance compared with state-of-the-art methods.