Deep Unitary Convolutional Neural Networks

Abstract

Deep neural networks can suffer from the exploding and vanishing activation problem, in which the networks fail to train properly because the neural signals either amplify or attenuate across the layers and become saturated. While other normalization methods aim to fix the stated problem, most of them have inference speed penalties in those applications that require running averages of the neural activations. Here we extend the unitary framework based on Lie algebra to neural networks of any dimensionalities, overcoming the major constraints of the prior arts that limit synaptic weights to be square matrices. Our proposed unitary convolutional neural networks deliver up to 32% faster inference speeds and up to 50% reduction in permanent hard disk space while maintaining competitive prediction accuracy.

References in corpus (10)

• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• On the difficulty of training Recurrent Neural Networks
• Instance Normalization: The Missing Ingredient for Fast Stylization
• On the Number of Linear Regions of Deep Neural Networks
• On orthogonality and learning recurrent networks with long term dependencies
• Full-Capacity Unitary Recurrent Neural Networks
• Orthogonal Weight Normalization: Solution to Optimization over Multiple Dependent Stiefel Manifolds in Deep Neural Networks
• Tunable Efficient Unitary Neural Networks (EUNN) and their application to RNNs
• Cheap Orthogonal Constraints in Neural Networks: A Simple Parametrization of the Orthogonal and Unitary Group
• Online Normalization for Training Neural Networks