Non-Volatile Reconfigurable Optical Digital Diffractive Neural Network Based on Phase Change Material

被引:0
作者
Hu, Qiaomu [1 ,2 ]
Zhao, Jingyu [1 ,2 ]
Wu, Chu [1 ,2 ]
Zeng, Rui [1 ,2 ]
Zhou, Xiaobing [1 ,2 ]
Zheng, Shuang [1 ,2 ,3 ]
Zhang, Minming [1 ,2 ,3 ]
机构
[1] Sch Opt & Elect Informat, Wuhan 430074, Peoples R China
[2] Natl Engn Res Ctr Next Generat Internet Access Sys, Wuhan 430074, Peoples R China
[3] Hubei Opt Fundamental Res Ctr, Wuhan 430074, Peoples R China
来源
IEEE PHOTONICS JOURNAL | 2024年 / 16卷 / 06期
基金
中国国家自然科学基金;
关键词
Adaptive optics; Neurons; Optical diffraction; Optical imaging; Biological neural networks; Optical computing; Photodetectors; Optical transmitters; Integrated optics; Phase change materials; Optical neural network; silicon photonics; phase change material;
D O I
10.1109/JPHOT.2024.3508052
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
Optical diffractive neural networks have sparked extensive research due to their low power consumption and high-speed capabilities in image processing. Here we propose and design a reconfigurable all-optical diffractive neural network structure with digital non-volatile optical neurons. The optical neurons are built with Sb2Se3 phase-change material and can switch between crystalline and amorphous states with no constant energy supply. Using three reconfigurable non-volatile digital diffractive layers and 10 photodetectors connected to a reconfigurable resistor network, our model achieves an accuracy of 94.46% in the handwritten digit recognition task. Moreover, the fabrication and assembly robustness of the proposed optical diffractive neural network is verified through full-vector diffractive simulation. Thanks to its reconfigurability and low energy supply, the digital optical diffractive neural network holds great potential to facilitate a programmable and low-power-consumption photonic processor for optical-artificial-intelligence.
引用
收藏
页数:8
相关论文
共 39 条
[1]   An overview on state-of-the-art electrocardiogram signal processing methods: Traditional to AI-based approaches [J].
Ardeti, Venkata Anuhya ;
Kolluru, Venkata Ratnam ;
Varghese, George Tom ;
Patjoshi, Rajesh Kumar .
EXPERT SYSTEMS WITH APPLICATIONS, 2023, 217
[2]   Reinforcement learning in a large-scale photonic recurrent neural network [J].
Bueno, J. ;
Maktoobi, S. ;
Froehly, L. ;
Fischer, I. ;
Jacquot, M. ;
Larger, L. ;
Brunner, D. .
OPTICA, 2018, 5 (06) :756-760
[3]   Hybrid optical-electronic convolutional neural networks with optimized diffractive optics for image classification [J].
Chang, Julie ;
Sitzmann, Vincent ;
Dun, Xiong ;
Heidrich, Wolfgang ;
Wetzstein, Gordon .
SCIENTIFIC REPORTS, 2018, 8
[4]   Pixelated non-volatile programmable photonic integrated circuits with 20-level intermediate states [J].
Chen, Wenyu ;
Liu, Shiyuan ;
Zhu, Jinlong .
INTERNATIONAL JOURNAL OF EXTREME MANUFACTURING, 2024, 6 (03)
[5]  
Cho K, 2014, ARXIV14061078, V2014, P1724, DOI DOI 10.3115/V1/D14-1179
[6]  
Courbariaux M, 2016, Arxiv, DOI [arXiv:1602.02830, DOI 10.48550/ARXIV.1602.02830, 10.48550/arXiv.1602.02830]
[7]   Nonvolatile programmable silicon photonics using an ultralow-loss Sb2Se3 phase change material [J].
Delaney, Matthew ;
Zeimpekis, Ioannis ;
Du, Han ;
Yan, Xingzhao ;
Banakar, Mehdi ;
Thomson, David J. ;
Hewak, Daniel W. ;
Muskens, Otto L. .
SCIENCE ADVANCES, 2021, 7 (25)
[8]   A New Family of Ultralow Loss Reversible Phase-Change Materials for Photonic Integrated Circuits: Sb2S3 and Sb2Se3 [J].
Delaney, Matthew ;
Zeimpekis, Ioannis ;
Lawson, Daniel ;
Hewak, Daniel W. ;
Muskens, Otto L. .
ADVANCED FUNCTIONAL MATERIALS, 2020, 30 (36)
[9]   All-optical spiking neurosynaptic networks with self-learning capabilities [J].
Feldmann, J. ;
Youngblood, N. ;
Wright, C. D. ;
Bhaskaran, H. ;
Pernice, W. H. P. .
NATURE, 2019, 569 (7755) :208-+
[10]   Reconfigurable binary diffractive optical neural network based on chalcogenide phase change material Ge2Sb2Se4Te1 [J].
Fu, Ziwei ;
Fu, Tingzhao ;
Wu, Hao ;
Zhu, Zhihong ;
Zhang, Jianfa .
OPTICS EXPRESS, 2024, 32 (23) :41433-41444