Metasurface-generated large and arbitrary analog convolution kernels for accelerated machine vision

TL;DR

This paper introduces a novel optical metasurface-based analog convolution method that significantly accelerates machine vision tasks by enabling large, arbitrary convolution kernels with high accuracy, surpassing digital methods in speed and efficiency.

Contribution

It presents a spatial frequency domain training approach to create customizable optical convolution kernels using metasurfaces, demonstrating high classification accuracy on standard datasets.

Findings

Achieved 98.59% accuracy on MNIST

Simulations show 92.63% on Fashion-MNIST

Demonstrated large, arbitrary kernels surpassing digital counterparts

Abstract

In the rapidly evolving field of artificial intelligence, convolutional neural networks are essential for tackling complex challenges such as machine vision and medical diagnosis. Recently, to address the challenges in processing speed and power consumption of conventional digital convolution operations, many optical components have been suggested to replace the digital convolution layer in the neural network, accelerating various machine vision tasks. Nonetheless, the analog nature of the optical convolution kernel has not been fully explored. Here, we develop a spatial frequency domain training method to create arbitrarily shaped analog convolution kernels using an optical metasurface as the convolution layer, with its receptive field largely surpassing digital convolution kernels. By employing spatial multiplexing, the multiple parallel convolution kernels with both positive and negative weights are generated under the incoherent illumination condition. We experimentally demonstrate a 98.59% classification accuracy on the MNIST dataset, with simulations showing 92.63% and 68.67% accuracy on the Fashion-MNIST and CIFAR-10 datasets with additional digital layers. This work underscores the unique advantage of analog optical convolution, offering a promising avenue to accelerate machine vision tasks, especially in edge devices.