一种脉冲卷积神经网络VLSI硬件架构设计

薛天志; 刘百成; 陈松

Chinese

您当前的位置：

首页 >

文章列表页 >

一种脉冲卷积神经网络VLSI硬件架构设计

更新时间：2026-09-04

- 一种脉冲卷积神经网络VLSI硬件架构设计
- A hardware implementation of a spiking convolutional neural network
- 微电子学与计算机 2019年36卷第10期页码：37-41
- 作者机构：
  
  中国科学技术大学,安徽,合肥,230026
- 作者简介：
- 基金信息：
  
  国家自然科学基金 61732020
- DOI：
  中图分类号：
- 纸质出版：2019
- 稿件说明：
移动端阅览
薛天志, 刘百成, 陈松. 一种脉冲卷积神经网络VLSI硬件架构设计[J]. 微电子学与计算机, 2019, 36(10): 37-41.

XUE Tian-zhi, LIU Bai-cheng, CHEN Song. A hardware implementation of a spiking convolutional neural network[J]. 2019, 36(10): 37-41.
薛天志, 刘百成, 陈松. 一种脉冲卷积神经网络VLSI硬件架构设计[J]. 微电子学与计算机, 2019, 36(10): 37-41. DOI：

XUE Tian-zhi, LIU Bai-cheng, CHEN Song. A hardware implementation of a spiking convolutional neural network[J]. 2019, 36(10): 37-41. DOI：

摘要

本文设计了一种识别手写体数字的脉冲卷积神经网络数字电路

使用脉冲神经元代替卷积核

并分别对卷积层和池化层设计相应的电路结构

实现全流水线并行.相比于传统的卷积神经网络

在识别MNIST数据集时

卷积神经网络的精确度为98.61%时

脉冲卷积神经网络的精度能达到98.04%.与相同流水线结构的卷积神经网络相比

脉冲神经网络平均能耗减少约50%.

Abstract

This paper designed a digital circuit of spiking convolutional neural network for recognizing handwriting number. Spiking neurons are used to replace the convolution kernels in CNN. Corresponding circuit structures are designed for convolution layer and pooling layer respectively to achieve full pipeline parallelism. Compared with the traditional convolutional neural network

when recognizing the MNIST dataset

the accuracy of the two is 98.61% and 98.04% respectively. Compared with CNN that has similar pipeline architecture

spiking convolutional neural network has a 50% reduction in average energy consumption.

关键词

Keywords

references

CAO Y, CHEN Y, KHOSLA D. Spiking deep convolutional neural networks for energy-efficient object recognition[J]. International Journal of Computer Vision, 2015, 113(1): 54-66. DOI: 54

DIEHL P U, NEIL D, BINAS J, et al. Fast-classifying, high-accuracy spiking deep networks through weight and threshold balancing[C]//2015 International Joint Conference on Neural Networks (IJCNN). IEEE, 2015: 1-8.

JIA Y, SHELHAMER E, DONAHUE J, et al. Caffe: Convolutional architecture for fast feature embedding[C]//Proceedings of the 22nd ACM international conference on Multimedia. ACM, 2014: 675-678.

ANWAR S, HWANG K, SUNG W. Fixed point optimization of deep convolutional neural networks for object recognition[C]//2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2015: 1131-1135.

LECUN Y, JACKEL L D, BOTTOU L, et al. Learning algorithms for classification: A comparison on handwritten digit recognition[J]. Neural networks: the statistical mechanics perspective, 1995, 261: 276. DOI: 276

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

V型沟槽SiC中子探测器的蒙特卡洛分析

基于MSS-YOLO-MPHands融合架构的装配手部动作识别研究

基于改进RT-DETR的航拍小目标检测算法

超低功耗实时时钟电路的设计

一种用于极低压boost系统的高效率电荷泵