期刊
IEEE TRANSACTIONS ON ELECTRON DEVICES
卷 66, 期 1, 页码 793-801出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TED.2018.2882779
关键词
Convolutional neural networks (CNNs); image recognition; online learning; resistive random-access memory (RRAM); synapse
资金
- NSFC [61334007, 61421005]
- National Innovation Training Program
- Beijing Municipal Science and Technology Plan Projects
In this paper, we devise and optimize schemes for the resistive random-access memory (RRAM)-based hardware implementation of convolutional neural networks (CNNs). The key achievements are as follows: 1) a specific circuit for CNN and corresponding operation methods is presented; 2) quantization schemes for utilizing binary RRAM or RRAM with multilevel resistances as synapses are proposed, and simulations show that our CNN system percentage accuracy for the MNIST data set using multilevel RRAM and 97% accuracy using binary RRAM as synapses; 3) the influence of factors such as the number and size of kernels, as well as the device conductance variation, on the final recognition accuracy is discussed. Three ways to reduce the hardware cost are also analyzed; and 4) we implement the online learning function using the developed CNN system with a binary spike-time-dependent plasticity protocol and achieve up to 94% accuracy for the online learning tasks.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据