Reduced storage direct tensor ring decomposition for convolutional neural networks compression

TL;DR

This paper introduces RSDTR, a novel tensor ring decomposition method for CNN compression that achieves high parameter and FLOPS reduction while maintaining accuracy, demonstrated on CIFAR-10 and ImageNet datasets.

Contribution

The paper proposes a new low-rank tensor ring decomposition method with increased permutation flexibility for CNN compression, outperforming existing approaches.

Findings

High compression rates achieved with minimal accuracy loss

Demonstrated effectiveness on CIFAR-10 and ImageNet datasets

Outperforms state-of-the-art CNN compression methods

Abstract

Convolutional neural networks (CNNs) are among the most widely used machine learning models for computer vision tasks, such as image classification. To improve the efficiency of CNNs, many CNNs compressing approaches have been developed. Low-rank methods approximate the original convolutional kernel with a sequence of smaller convolutional kernels, which leads to reduced storage and time complexities. In this study, we propose a novel low-rank CNNs compression method that is based on reduced storage direct tensor ring decomposition (RSDTR). The proposed method offers a higher circular mode permutation flexibility, and it is characterized by large parameter and FLOPS compression rates, while preserving a good classification accuracy of the compressed network. The experiments, performed on the CIFAR-10 and ImageNet datasets, clearly demonstrate the efficiency of RSDTR in comparison to other state-of-the-art CNNs compression approaches.