Neural network pruning method and device, equipment and storage medium

Abstract

The invention discloses a neural network pruning method and device, equipment and a storage medium. When the neural network is pruned, a network layer to be pruned can be used as a target network layer for channel pruning and convolution kernel reconstruction. Therefore, when the multi-branch structure is compressed by the scheme, the compression is not limited to an intermediate layer; network layers such as an input layer, an output layer, a down-sampling layer and the like can be compressed, so that the compression ratio of the neural network is greatly improved, the calculation amount required by a neural network model to execute tasks is reduced, and the task processing speed of the neural network is increased; moreover, the scheme is a data-independent asynchronous channel pruning method, which is beneficial to keeping the robustness of the compressed neural network, and can also realize pruning of different sparse granularities in different network layers of the neural network, thereby improving the compression flexibility.

Description

technical field [0001] The present invention relates to the technical field of deep neural network compression and acceleration, and more specifically, to a neural network pruning method, device, equipment and storage medium. Background technique [0002] As the depth and width of the neural network increase, it shows excellent performance in various AI (Artificial Intelligence, artificial intelligence) application scenarios, such as machine vision tasks such as image recognition and target detection. Faced with the development trend of machine vision software based on deep learning in various embedded devices or mobile devices, deep neural networks with huge parameters are difficult to deploy in devices with limited computing and storage resources. Deep neural network compression and acceleration technology provides a solution for the long-term real-time application of deep learning in these resource-constrained devices. The deep neural network compression technology achie...

AI Technical Summary

Problems solved by technology

[0003] At present, the convolutional neural network has evolved a variety of multi-branch structures, but the pruning schemes for multi-branch structures only cut the input and output channels of the middle layer of the bottleneck structure. For the input and output of the entire multi-branch structure The channel is not compressed, and the number of channels in the middle layer of the bottleneck structure of the multi-branch structure is originally smaller than the number of input and output channels of the entire module. Therefore, only compressing the middle layer of the multi-branch structure is not conducive to improving the compression ratio of the multi-branch structure.

Images