A Frobenius norm regularization method for convolutional kernels to avoid unstable gradient problem

Abstract

Convolutional neural network is a very important model of deep learning. It can help avoid the exploding/vanishing gradient problem and improve the generalizability of a neural network if the singular values of the Jacobian of a layer are bounded around $1$ in the training process. We propose a new penalty function for a convolutional kernel to let the singular values of the corresponding transformation matrix are bounded around $1$. We show how to carry out the gradient type methods. The penalty is about the structured transformation matrix corresponding to a convolutional kernel. This provides a new regularization method about the weights of convolutional layers.

arXiv commentarXiv admin note: text overlap with arXiv:1906.04866

References in corpus (5)

• Spectral Normalization for Generative Adversarial Networks
• A guide to convolution arithmetic for deep learning
• Spectral Norm Regularization for Improving the Generalizability of Deep Learning
• Lipschitz-Margin Training: Scalable Certification of Perturbation Invariance for Deep Neural Networks
• The Singular Values of Convolutional Layers