An Optical Frontend for a Convolutional Neural Network

TL;DR

This paper presents a hybrid optical-electronic neural network architecture that leverages nanophotonic components for efficient linear operations, reducing power consumption and increasing speed for large-scale image processing tasks.

Contribution

It introduces a novel optical frontend for the first layer of a CNN, minimizing electronic-optical conversions and enhancing performance over purely electronic systems.

Findings

Outperforms electronic architectures in speed and power for large images and kernels.

Achieves 87% accuracy on Kaggle Cats and Dogs dataset with modified AlexNet.

Design demonstrates feasibility of hybrid photonic-electronic CNNs.

Abstract

The parallelism of optics and the miniaturization of optical components using nanophotonic structures, such as metasurfaces present a compelling alternative to electronic implementations of convolutional neural networks. The lack of a low-power optical nonlinearity, however, requires slow and energy-inefficient conversions between the electronic and optical domains. Here, we design an architecture which utilizes a single electrical to optical conversion by designing a free-space optical frontend unit that implements the linear operations of the first layer with the subsequent layers realized electronically. Speed and power analysis of the architecture indicates that the hybrid photonic-electronic architecture outperforms sole electronic architecture for large image sizes and kernels. Benchmarking of the photonic-electronic architecture on a modified version of AlexNet achieves a classification accuracy of 87% on images from the Kaggle Cats and Dogs challenge database.