Abstract
Stochasticity is an inherent feature of biological neural activities. We propose a noise-injection scheme to implement a GHz-rate stochastic photonic spiking neuron (S-PSN). The firing-probability encoding is experimentally demonstrated and exploited for Bayesian inference with unsupervised learning. In a breast diagnosis task, the stochastic photonic spiking neural network (S-PSNN) can not only achieve a classification accuracy of 96.6%, but can also evaluate the diagnosis uncertainty with prediction entropies. As a result, the misdiagnosis rate is reduced by 80% compared to that of a conventional deterministic photonic spiking neural network (D-PSNN) for the same task. The GHz-rate S-PSN endows the neuromorphic photonics with high-speed Bayesian inference for reliable information processing in error-critical scenarios.
© 2023 Optica Publishing Group
Full Article | PDF ArticleMore Like This
Shuiying Xiang, Yuechun Shi, Xingxing Guo, Yahui Zhang, Hongji Wang, Dianzhuang Zheng, Ziwei Song, Yanan Han, Shuang Gao, Shihao Zhao, Biling Gu, Hailing Wang, Xiaojun Zhu, Lianping Hou, Xiangfei Chen, Wanhua Zheng, Xiaohua Ma, and Yue Hao
Optica 10(2) 162-171 (2023)
Yun-Jhu Lee, Mehmet Berkay On, Xian Xiao, Roberto Proietti, and S. J. Ben Yoo
Opt. Express 30(11) 19360-19389 (2022)
Qiang Zhang, Ning Jiang, Anran Li, Yiqun Zhang, Gang Hu, Yongsheng Cao, and Kun Qiu
Opt. Lett. 49(8) 1965-1968 (2024)