Learning to Optimize Domain Specific Normalization for Domain Generalization

TL;DR

This paper introduces a multi-source domain generalization method that optimizes domain-specific normalization layers using multiple normalization techniques, improving model generalizability across diverse domains.

Contribution

It proposes a novel normalization approach combining multiple normalization methods with learned domain-specific parameters to enhance domain generalization performance.

Findings

Achieves state-of-the-art accuracy on standard benchmarks.

Effective in multi-source domain adaptation tasks.

Robust to label noise in domain generalization scenarios.

Abstract

We propose a simple but effective multi-source domain generalization technique based on deep neural networks by incorporating optimized normalization layers that are specific to individual domains. Our approach employs multiple normalization methods while learning separate affine parameters per domain. For each domain, the activations are normalized by a weighted average of multiple normalization statistics. The normalization statistics are kept track of separately for each normalization type if necessary. Specifically, we employ batch and instance normalizations in our implementation to identify the best combination of these two normalization methods in each domain. The optimized normalization layers are effective to enhance the generalizability of the learned model. We demonstrate the state-of-the-art accuracy of our algorithm in the standard domain generalization benchmarks, as well as viability to further tasks such as multi-source domain adaptation and domain generalization in the presence of label noise.