Online Normalization for Training Neural Networks
This addresses the problem of normalization in neural networks for researchers and practitioners, enabling use in cases like recurrent networks where batching is prohibitive, though it is incremental as it builds on existing normalization methods.
The paper introduces Online Normalization, a technique that normalizes hidden activations without using batches, achieving accuracy comparable to Batch Normalization while resolving its theoretical gradient limitation. It demonstrates applications in image classification, image segmentation, and language modeling with experimental results on datasets like ImageNet, CIFAR, and PTB.
Online Normalization is a new technique for normalizing the hidden activations of a neural network. Like Batch Normalization, it normalizes the sample dimension. While Online Normalization does not use batches, it is as accurate as Batch Normalization. We resolve a theoretical limitation of Batch Normalization by introducing an unbiased technique for computing the gradient of normalized activations. Online Normalization works with automatic differentiation by adding statistical normalization as a primitive. This technique can be used in cases not covered by some other normalizers, such as recurrent networks, fully connected networks, and networks with activation memory requirements prohibitive for batching. We show its applications to image classification, image segmentation, and language modeling. We present formal proofs and experimental results on ImageNet, CIFAR, and PTB datasets.