TaskNorm: Rethinking Batch Normalization for Meta-Learning

TL;DR

This paper introduces TaskNorm, a novel batch normalization method tailored for meta-learning, demonstrating significant improvements in accuracy and training efficiency across multiple datasets.

Contribution

The paper proposes TaskNorm, a new normalization approach specifically designed for meta-learning, addressing limitations of conventional batch normalization in hierarchical settings.

Findings

TaskNorm consistently improves meta-learning performance.

Normalization choice significantly impacts accuracy and training time.

Experiments on 14 datasets validate the effectiveness of TaskNorm.

Abstract

Modern meta-learning approaches for image classification rely on increasingly deep networks to achieve state-of-the-art performance, making batch normalization an essential component of meta-learning pipelines. However, the hierarchical nature of the meta-learning setting presents several challenges that can render conventional batch normalization ineffective, giving rise to the need to rethink normalization in this setting. We evaluate a range of approaches to batch normalization for meta-learning scenarios, and develop a novel approach that we call TaskNorm. Experiments on fourteen datasets demonstrate that the choice of batch normalization has a dramatic effect on both classification accuracy and training time for both gradient based and gradient-free meta-learning approaches. Importantly, TaskNorm is found to consistently improve performance. Finally, we provide a set of best practices for normalization that will allow fair comparison of meta-learning algorithms.