Conversion of a single-layer ANN to photonic SNN for pattern recognition

作     者：Yanan HAN Shuiying XIANG Tianrui ZHANG Yahui ZHANG Xingxing GUO Yuechun SHI 

作者机构：State Key Laboratory of Integrated Service NetworksState Key Discipline Laboratory of Wide Bandgap Semiconductor TechnologyXidian University Yongjiang Laboratory 

出 版 物：《Science China(Information Sciences)》 (中国科学:信息科学(英文版))

年 卷 期：2024年第67卷第1期

页      面：261-270页

核心收录：

学科分类：12[管理学] 1201[管理学-管理科学与工程(可授管理学、工学学位)] 081104[工学-模式识别与智能系统] 08[工学] 0835[工学-软件工程] 0811[工学-控制科学与工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：supported in part by National Key Research and Development Program of China (Grant Nos.2021YFB2801900,2021YFB2801901,2021YFB2801902,2021YFB2801904) National Natural Science Foundation of China(Grant Nos.61974177,61674119) National Outstanding Youth Science Fund Project of National Natural Science Foundation of China (Grant No.62022062) Fundamental Research Funds for the Central Universities (Grant No.JB210114) 

主　　题：photonic SNN conversion optical computing pattern recognition artificial neural network 

摘      要：This work presents a complete conversion scheme for photonic spiking neural networks(SNNs).We verified that the output of an artificial neural network(ANN) trained with the simulated optical activation function can be directly converted into the spike rate of a photonic spiking neuron *** reveal the feasibility of hardware implementation,we considered the effects of different bit precisions of data and weight,noise level,and bias current mismatch on the converted *** proposed scheme was evaluated using the Deterding vowel,IRIS,TIDIGITS,and MNIST datasets for pattern recognition,and achieved mean accuracies of 95.80%,98.67%,96.19%,and 92.33%,*** proposed scheme can convert an ANN into a photonic SNN with almost no precision loss,and the performance was comparable to that of an ANN trained with the rectified linear unit *** proposed scheme can enable the high-performance implementation of photonic SNNs.