Pseudo Multi-Source Domain Generalization: Bridging the Gap Between Single and Multi-Source Domain Generalization

Shohei Enomoto

Keywords: Domain Generalization, Data Augmentation, Style Transformation, Computer Vision, Deep Learning, Benchmark

TL;DR: We propose a new framework that generates pseudo multi-domain datasets and enables the application of multi-domain generalization algorithms to single-domain datasets.

Abstract: Deep learning models often struggle to maintain performance when deployed on data distributions different from their training data, particularly in real-world applications where environmental conditions frequently change. While multi-source domain generalization (MDG) has shown promise in addressing this challenge by leveraging multiple source domains during training, its practical application is limited by the significant costs and difficulties associated with creating multi-domain datasets. To address this limitation, we propose pseudo multi-source domain generalization (PMDG), a novel framework that enables the application of sophisticated MDG algorithms in a more practical single-source domain generalization setting. PMDG generates multiple pseudo-domains from a single source domain through style transfer and data augmentation techniques, creating a synthetic multi-domain dataset that can be used with MDG algorithms. Through extensive experiments with PseudoDomainBed, our modified version of the DomainBed benchmark, we analyze the effectiveness of PMDG across multiple datasets and architectures. Our analysis reveals several key findings, including a positive correlation between MDG and PMDG performance and the potential of pseudo-domains to match or exceed actual multi-domain performance with sufficient data. These comprehensive empirical results provide valuable insights for future research in domain generalization.

Primary Area: other topics in machine learning (i.e., none of the above)

Submission Number: 7211