Normalizing the Normalizers: Comparing and Extending Network Normalization Schemes

TL;DR

This paper presents a unified divisive normalization framework that generalizes existing normalization techniques and introduces modifications with regularization, leading to improved performance across various neural network architectures and tasks.

Contribution

It unifies batch and layer normalization as divisive normalization and proposes modifications with regularization that enhance model performance.

Findings

Improved accuracy in image classification, language modeling, and super-resolution.

Unified framework simplifies understanding of normalization techniques.

Modified normalization schemes outperform standard methods.

Abstract

Normalization techniques have only recently begun to be exploited in supervised learning tasks. Batch normalization exploits mini-batch statistics to normalize the activations. This was shown to speed up training and result in better models. However its success has been very limited when dealing with recurrent neural networks. On the other hand, layer normalization normalizes the activations across all activities within a layer. This was shown to work well in the recurrent setting. In this paper we propose a unified view of normalization techniques, as forms of divisive normalization, which includes layer and batch normalization as special cases. Our second contribution is the finding that a small modification to these normalization schemes, in conjunction with a sparse regularizer on the activations, leads to significant benefits over standard normalization techniques. We demonstrate the effectiveness of our unified divisive normalization framework in the context of convolutional neural nets and recurrent neural networks, showing improvements over baselines in image classification, language modeling as well as super-resolution.