Tensor decomposition-based acceleration and compression method for deep convolutional neural network

Abstract

The invention discloses a tensor decomposition-based acceleration and compression method for a deep convolutional neural network. The method at least comprises the steps of 1: obtaining an original deep convolutional neural network; 2: performing tensor decomposition on a weight tensor of each layer in the original deep convolutional neural network to obtain a plurality of low-rank sub-tensors; and 3: replacing the weight tensor of each layer in the original deep convolutional neural network with the low-rank sub-tensors to obtain a new deep convolutional neural network. Through an embodiment of the method, the acceleration and compression of a large deep convolutional neural network are realized.

Description

technical field [0001] Embodiments of the present invention relate to the technical field of deep neural networks, and in particular, to a method for accelerating and compressing deep convolutional neural networks based on tensor decomposition. Background technique [0002] In recent years, deep convolutional neural networks have made great breakthroughs in many fields such as computer vision and speech processing, significantly improving the performance of tasks such as image retrieval and classification, object detection, object tracking, and speech recognition. It has been widely used in many industries such as video surveillance, entertainment, and smart home. [0003] The breakthrough of deep convolutional neural networks is largely due to new computing tools, such as GPUs, computer clusters, etc., as well as large-scale datasets containing manually labeled information. On the other hand, the number of layers of convolutional neural networks has been increasing in rece...

AI Technical Summary

Problems solved by technology

Simultaneous acceleration and compression of all layers o...

Images