Nonlinear Processing with Linear Optics

TL;DR

This paper introduces a novel optical framework utilizing multiple scattering to perform programmable linear and nonlinear transformations at low power, aiming to improve energy efficiency in neural network implementations.

Contribution

The study presents a new method for optical neural networks that combines multiple scattering with nonlinear transformations without electronic components.

Findings

Enables nonlinear optical computing at low power using multiple scattering.

Experimental and theoretical validation of the framework's effectiveness.

Scaling behavior follows the power law similar to digital neural networks.

Abstract

Deep neural networks have achieved remarkable breakthroughs by leveraging multiple layers of data processing to extract hidden representations, albeit at the cost of large electronic computing power. To enhance energy efficiency and speed, the optical implementation of neural networks aims to harness the advantages of optical bandwidth and the energy efficiency of optical interconnections. In the absence of low-power optical nonlinearities, the challenge in the implementation of multilayer optical networks lies in realizing multiple optical layers without resorting to electronic components. In this study, we present a novel framework that uses multiple scattering that is capable of synthesizing programmable linear and nonlinear transformations concurrently at low optical power by leveraging the nonlinear relationship between the scattering potential, represented by data, and the scattered field. Theoretical and experimental investigations show that repeating the data by multiple scattering enables non-linear optical computing at low power continuous wave light. Moreover, we empirically found that scaling of this optical framework follows the power law as in state-of-the-art deep digital networks.