Abstract

The existing implementations of reconfigurable diffractive neural networks rely on both a liquid-crystal spatial light modulator and a digital micromirror device, which results in complexity in the alignment of the optical system and a constrained computational speed. Here, we propose a superpixel diffractive neural network that leverages solely a digital micromirror device to control the neuron bias and connection. This approach considerably simplifies the optical system and achieves a computational speed of 326 Hz per neural layer. We validate our method through experiments in digit classification, achieving an accuracy of 82.6%, and action recognition, attaining a perfect accuracy of 100%. Our findings demonstrate the effectiveness of the superpixel diffractive neural network in simplifying the optical system and enhancing computational speed, opening up new possibilities for real-time optical information processing applications.

© 2023 Optica Publishing Group

Effects of interlayer reflection and interpixel interaction in diffractive optical neural networks

Minhan Lou, Yingjie Li, Cunxi Yu, Berardi Sensale-Rodriguez, and Weilu Gao
Opt. Lett. 48(2) 219-222 (2023)

High-definition image transmission through dynamically perturbed multimode fiber by a self-attention based neural network

Guohua Wu, Yong Sun, Longfei Yin, Zhixiong Song, and Wenting Yu
Opt. Lett. 48(10) 2764-2767 (2023)

Physics-model-based neural networks for inverse design of binary phase planar diffractive lenses

Jianmin He, Zhenghao Guo, Yongying Zhang, Yiyang Lu, Feng Wen, Haixia Da, Guofu Zhou, Dong Yuan, and Huapeng Ye
Opt. Lett. 48(6) 1474-1477 (2023)

Orthogonality of diffractive deep neural network

Shuiqin Zheng, Shixiang Xu, and Dianyuan Fan
Opt. Lett. 47(7) 1798-1801 (2022)

Convolutional-neural-network-based versus vision-transformer-based SNR estimation for visible light communication networks

Zhenquan Zhao, Faisal Nadeem Khan, Zeyad A. H. Qasem, Bohua Deng, Qian Li, Zhengyong Liu, and H. Y. Fu
Opt. Lett. 48(6) 1419-1422 (2023)