2024-03-28T08:59:00Z
https://nagoya.repo.nii.ac.jp/oai
oai:nagoya.repo.nii.ac.jp:02002704
2023-01-16T05:09:12Z
312:313:314
Neural Network Calculations at the Speed of Light Using Optical Vector-Matrix Multiplication and Optoelectronic Activation
HATTORI, Naoki
SHIOMI, Jun
MASUDA, Yutaka
ISHIHARA, Tohru
SHINYA, Akihiko
NOTOMI, Masaya
open access
Copyright(C)2021 IEICE
neural network
optical circuit
multi-layer perceptron
wavelength division multiplexing
With the rapid progress of the integrated nanophotonics technology, the optical neural network architecture has been widely investigated. Since the optical neural network can complete the inference processing just by propagating the optical signal in the network, it is expected more than one order of magnitude faster than the electronics-only implementation of artificial neural networks (ANN). In this paper, we first propose an optical vector-matrix multiplication (VMM) circuit using wavelength division multiplexing, which enables inference processing at the speed of light with ultra-wideband. This paper next proposes optoelectronic circuit implementation for batch normalization and activation function, which significantly improves the accuracy of the inference processing without sacrificing the speed performance. Finally, using a virtual environment for machine learning and an optoelectronic circuit simulator, we demonstrate the ultra-fast and accurate operation of the optical-electronic ANN circuit.
電子情報通信学会
2022-05-11
2021-11-01
eng
journal article
VoR
http://hdl.handle.net/2237/0002002704
https://nagoya.repo.nii.ac.jp/records/2002704
https://doi.org/10.1587/transfun.2020KEP0016
0916-8508
IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences
E104A
11
1477
1487
https://nagoya.repo.nii.ac.jp/record/2002704/files/e104-a_11_1477.pdf
application/pdf
4.4 MB
2022-05-11