Generative appearance replay for continual unsupervised domain adaptation

TL;DR

This paper introduces GarDA, a generative replay method for continual unsupervised domain adaptation in medical image segmentation, enabling models to adapt sequentially to multiple domains without retaining past data.

Contribution

GarDA is the first approach to perform continual unsupervised domain adaptation for segmentation without needing access to previous data, addressing privacy and storage constraints.

Findings

GarDA outperforms existing methods on multiple datasets.

It effectively consolidates information across multiple domains.

The approach is applicable in privacy-sensitive scenarios.

Abstract

Deep learning models can achieve high accuracy when trained on large amounts of labeled data. However, real-world scenarios often involve several challenges: Training data may become available in installments, may originate from multiple different domains, and may not contain labels for training. Certain settings, for instance medical applications, often involve further restrictions that prohibit retention of previously seen data due to privacy regulations. In this work, to address such challenges, we study unsupervised segmentation in continual learning scenarios that involve domain shift. To that end, we introduce GarDA (Generative Appearance Replay for continual Domain Adaptation), a generative-replay based approach that can adapt a segmentation model sequentially to new domains with unlabeled data. In contrast to single-step unsupervised domain adaptation (UDA), continual adaptation to a sequence of domains enables leveraging and consolidation of information from multiple domains. Unlike previous approaches in incremental UDA, our method does not require access to previously seen data, making it applicable in many practical scenarios. We evaluate GarDA on two datasets with different organs and modalities, where it substantially outperforms existing techniques.