$L_2$BN: Enhancing Batch Normalization by Equalizing the $L_2$ Norms of Features

TL;DR

This paper introduces $L_2$BN, a simple method that normalizes feature magnitudes before batch normalization to improve class discriminability and model generalization in image and acoustic classification tasks.

Contribution

The paper proposes $L_2$BN, a novel normalization technique that enhances batch normalization by equalizing feature norms without extra parameters.

Findings

Improves intra-class compactness of features.

Increases inter-class feature discrepancy.

Boosts generalization across models and tasks.

Abstract

In this paper, we analyze batch normalization from the perspective of discriminability and find the disadvantages ignored by previous studies: the difference in $l_2$ norms of sample features can hinder batch normalization from obtaining more distinguished inter-class features and more compact intra-class features. To address this issue, we propose a simple yet effective method to equalize the $l_2$ norms of sample features. Concretely, we $l_2$-normalize each sample feature before feeding them into batch normalization, and therefore the features are of the same magnitude. Since the proposed method combines the $l_2$ normalization and batch normalization, we name our method $L_2$BN. The $L_2$BN can strengthen the compactness of intra-class features and enlarge the discrepancy of inter-class features. The $L_2$BN is easy to implement and can exert its effect without any additional parameters or hyper-parameters. We evaluate the effectiveness of $L_2$BN through extensive experiments with various models on image classification and acoustic scene classification tasks. The results demonstrate that the $L_2$BN can boost the generalization ability of various neural network models and achieve considerable performance improvements.