Differentiable Learning-to-Normalize via Switchable Normalization

TL;DR

Switchable Normalization (SN) dynamically learns to select the most suitable normalization method for each layer in deep neural networks, improving robustness and performance across various architectures, tasks, and batch sizes.

Contribution

SN introduces a learnable normalization method that combines multiple normalization techniques, enhancing adaptability and robustness in deep learning models.

Findings

SN outperforms existing normalization methods on benchmarks like ImageNet and COCO.

SN maintains high performance even with small batch sizes.

SN is less sensitive to hyper-parameters compared to other normalization techniques.

Abstract

We address a learning-to-normalize problem by proposing Switchable Normalization (SN), which learns to select different normalizers for different normalization layers of a deep neural network. SN employs three distinct scopes to compute statistics (means and variances) including a channel, a layer, and a minibatch. SN switches between them by learning their importance weights in an end-to-end manner. It has several good properties. First, it adapts to various network architectures and tasks (see Fig.1). Second, it is robust to a wide range of batch sizes, maintaining high performance even when small minibatch is presented (e.g. 2 images/GPU). Third, SN does not have sensitive hyper-parameter, unlike group normalization that searches the number of groups as a hyper-parameter. Without bells and whistles, SN outperforms its counterparts on various challenging benchmarks, such as ImageNet, COCO, CityScapes, ADE20K, and Kinetics. Analyses of SN are also presented. We hope SN will help ease the usage and understand the normalization techniques in deep learning. The code of SN has been made available in https://github.com/switchablenorms/.