Hybrid training of optical neural networks
Top Cited Papers
Open Access
- 14 July 2022
- journal article
- research article
- Published by Optica Publishing Group in Optica
- Vol. 9 (7) , 803-811
Abstract
Optical neural networks are emerging as a promising type of machine learning hardware capable of energy-efficient, parallel computation. Today’s optical neural networks are mainly developed to perform optical inference after in silico training on digital simulators. However, various physical imperfections that cannot be accurately modeled may lead to the notorious “reality gap” between the digital simulator and the physical system. To address this challenge, we demonstrate hybrid training of optical neural networks where the weight matrix is trained with neuron activation functions computed optically via forward propagation through the network. We examine the efficacy of hybrid training with three different networks: an optical linear classifier, a hybrid opto-electronic network, and a complex-valued optical network. We perform a study comparative to in silico training, and our results show that hybrid training is robust against different kinds of static noise. Our platform-agnostic hybrid training scheme can be applied to a wide variety of optical neural networks, and this work paves the way towards advanced all-optical training in machine intelligence.Keywords
All Related Versions
Funding Information
- Engineering and Physical Sciences Research Council (IAA)
- Royal Commission for the Exhibition of 1851
This publication has 34 references indexed in Scilit:
- Setting up meshes of interferometers – reversed local light interference methodOptics Express, 2017
- Attojoule Optoelectronics for Low-Energy Information Processing and CommunicationsJournal of Lightwave Technology, 2017
- Perfect optics with imperfect componentsOptica, 2015
- Deep learningNature, 2015
- Photonic information processing beyond Turing: an optoelectronic implementation of reservoir computingOptics Express, 2012
- Pixelated phase computer holograms for the accurate encoding of scalar complex fieldsJournal of the Optical Society of America A, 2007
- Reinforcement and backpropagation training for an optical neural network using self-lensing effectsIEEE Transactions on Neural Networks, 2000
- Optical Neural ComputersScientific American, 1987
- Optical implementation of the Hopfield modelApplied Optics, 1985
- Binary computer-generated hologramsApplied Optics, 1979