DATTA: Domain Diversity Aware Test-Time Adaptation for Dynamic Domain Shift Data Streams

TL;DR

DATTA introduces a novel test-time adaptation method that effectively handles dynamic domain shifts in data streams by recognizing domain diversity and adapting normalization and fine-tuning processes, significantly improving performance.

Contribution

It is the first method to address TTA under dynamic domain shifts by incorporating domain-diversity recognition and adaptive strategies for normalization and fine-tuning.

Findings

Outperforms state-of-the-art methods by up to 13%.

Effectively handles single- and multi-domain shifts in data streams.

Improves adaptation accuracy in dynamic domain scenarios.

Abstract

Test-Time Adaptation (TTA) addresses domain shifts between training and testing. However, existing methods assume a homogeneous target domain (e.g., single domain) at any given time. They fail to handle the dynamic nature of real-world data, where single-domain and multiple-domain distributions change over time. We identify that performance drops in multiple-domain scenarios are caused by batch normalization errors and gradient conflicts, which hinder adaptation. To solve these challenges, we propose Domain Diversity Adaptive Test-Time Adaptation (DATTA), the first approach to handle TTA under dynamic domain shift data streams. It is guided by a novel domain-diversity score. DATTA has three key components: a domain-diversity discriminator to recognize single- and multiple-domain patterns, domain-diversity adaptive batch normalization to combine source and test-time statistics, and domain-diversity adaptive fine-tuning to resolve gradient conflicts. Extensive experiments show that DATTA significantly outperforms state-of-the-art methods by up to 13%. Code is available at https://github.com/DYW77/DATTA.