Log In Sign Up

Backward Gradient Normalization in Deep Neural Networks

by   Alejandro Cabana, et al.

We introduce a new technique for gradient normalization during neural network training. The gradients are rescaled during the backward pass using normalization layers introduced at certain points within the network architecture. These normalization nodes do not affect forward activity propagation, but modify backpropagation equations to permit a well-scaled gradient flow that reaches the deepest network layers without experimenting vanishing or explosion. Results on tests with very deep neural networks show that the new technique can do an effective control of the gradient norm, allowing the update of weights in the deepest layers and improving network accuracy on several experimental conditions.


page 1

page 2

page 3

page 4


Analysis on Gradient Propagation in Batch Normalized Residual Networks

We conduct mathematical analysis on the effect of batch normalization (B...

Channel Normalization in Convolutional Neural Network avoids Vanishing Gradients

Normalization layers are widely used in deep neural networks to stabiliz...

Softmax Gradient Tampering: Decoupling the Backward Pass for Improved Fitting

We introduce Softmax Gradient Tampering, a technique for modifying the g...

Designing Network Design Strategies Through Gradient Path Analysis

Designing a high-efficiency and high-quality expressive network architec...

Volume-preserving Neural Networks: A Solution to the Vanishing Gradient Problem

We propose a novel approach to addressing the vanishing (or exploding) g...

Scale Normalization

One of the difficulties of training deep neural networks is caused by im...