Abstract:Training an artificial neural network with backpropagation algorithms to perform advanced machine learning tasks requires an extensive computational process. This paper proposes to implement the backpropagation algorithm optically for in situ training of both linear and nonlinear diffractive optical neural networks, which enables the acceleration of training speed and improvement in energy efficiency on core computing modules. We demonstrate that the gradient of a loss function with respect to the weights of diffractive layers can be accurately calculated by measuring the forward and backward propagated optical fields based on light reciprocity and phase conjunction principles. The diffractive modulation weights are updated by programming a high-speed spatial light modulator to minimize the error between prediction and target output and perform inference tasks at the speed of light. We numerically validate the effectiveness of our approach on simulated networks for various applications. The proposed in situ optical learning architecture achieves accuracy comparable to in silico training with an electronic computer on the tasks of object classification and matrix-vector multiplication, which further allows the diffractive optical neural network to adapt to system imperfections. Also, the self-adaptive property of our approach facilitates the novel application of the network for all-optical imaging through scattering media. The proposed approach paves the way for robust implementation of large-scale diffractive neural networks to perform distinctive tasks all-optically. (C) 2020 Chinese Laser Press

Optical Machine Learning with Incoherent Light and a Single-Pixel Detector

Optical Analog Computing Devices Designed by Deep Neural Network

Optical Machine Learning with Single-pixel Imaging

Digit Recognition Based on Programmable Nanophotonic Processor

Single-Pixel Machine Vision Using Spectral Encoding Through Diffractive Optical Networks

All-Optical Machine Learning Using Diffractive Deep Neural Networks

Deep learning with coherent VCSEL neural networks

Low-power scalable multilayer optoelectronic neural networks enabled with incoherent light

1-bit Quantized On-chip Hybrid Diffraction Neural Network Enabled by Authentic All-optical Fully-connected Architecture

All-optical Fourier neural network using partially coherent light

Space-efficient optical computing with an integrated chip diffractive neural network

A surface-normal photodetector as nonlinear activation function in diffractive optical neural networks

Fully forward mode training for optical neural networks

Optical-electronic neural network for multi-modality and high-accurate lensless optics design and image reconstruction

Implementation of Optical Deep Neural Networks using the Fabry-Perot Interferometer

Single-shot optical neural network

Machine learning meets Singular Optics: Speckle-based Structured light demultiplexing

Optical authentication scheme based on all-optical neural network

<i>In situ</i> optical backpropagation training of diffractive optical neural networks

Multimodal deep learning using on-chip diffractive optics with in situ training capability

Nanoprinted high-neuron-density optical linear perceptrons performing near-infrared inference on a CMOS chip