Memory-based convolutional neural network system

Abstract

The invention discloses a memory-based convolutional neural network system, comprising an input module, a convolutional layer circuit module, a pooling circuit module, an activation function module, afully connected layer circuit module and an output module, wherein a convolution kernel value or a synaptic weight value is stored in a NOR FLASH unit; the input module converts an input signal intoa voltage signal required by a convolutional neural network; the convolutional layer circuit module performs an convolution operation on the voltage signal corresponding to the input signal and the convolution kernel value and transmits the result to the activation function module; the activation function module activates the signal; the pooling layer circuit performs a pooling operation on the activated signal; and the fully connected layer circuit module multiplies the signal subjected to pooling operation with the synaptic weight value to perform classification, and a softmax function module normalizes the classification result to a probability value and takes the classification result as the output of the entire network. The system disclosed by the invention can satisfy the requirements of real-time data processing, and has a low hardware cost.

Description

[0001] This application claims the priority of the Chinese patent application submitted to the Patent Office of the State Intellectual Property Office of China on May 8, 2018, the application number is 201810434049.6, and the invention title is "A Convolutional Neural Network System Based on NOR FLASH", all of which The contents are incorporated by reference in this application. technical field [0002] The present invention relates to the technical field of artificial neural networks, and more specifically, to a memory-based convolutional neural network system. Background technique [0003] The network structure of Convolutional Neural Network (CNN) was first proposed by Fukushima in 1980. However, due to the difficulty of implementing the training algorithm, it has not been widely used. In the 1990s, LeCun et al. applied a gradient-based learning algorithm to CNN and achieved good results. Since then, researchers have further refined CNNs and achieved strong results in t...

AI Technical Summary

Problems solved by technology

[0005] Aiming at the defects of the prior art, the purpose of the present invention is to solve the problem that the existing deep convolutional neural network always needs a large amount of training data, and utilizes the weight parameter to reach 10 8 Level, getting data from the memory and sending it to the CPU and GPU for calculation and sending the result back to the storage will become quite time-consuming, which cannot meet the needs of real-time data processing, and may also cause technical problems with huge hardware costs

Images