High-performance real-world optical computing trained by in situ gradient-based model-free optimization

TL;DR

This paper introduces a gradient-based, model-free optimization method for optical computing systems that enables efficient in situ training without heavy simulations, improving performance on real-world tasks.

Contribution

The authors develop G-MFO, a Monte Carlo gradient estimation approach that trains optical systems directly as black boxes, bypassing complex simulations and enabling practical deployment.

Findings

G-MFO outperforms hybrid training on MNIST and FMNIST datasets.

It enables high-speed, image-free classification of cells from phase maps.

The method requires low computational resources and is suitable for real-world applications.

Abstract

Optical computing systems provide high-speed and low-energy data processing but face deficiencies in computationally demanding training and simulation-to-reality gaps. We propose a gradient-based model-free optimization (G-MFO) method based on a Monte Carlo gradient estimation algorithm for computationally efficient in situ training of optical computing systems. This approach treats an optical computing system as a black box and back-propagates the loss directly to the optical computing weights' probability distributions, circumventing the need for a computationally heavy and biased system simulation. Our experiments on diffractive optical computing systems show that G-MFO outperforms hybrid training on the MNIST and FMNIST datasets. Furthermore, we demonstrate image-free and high-speed classification of cells from their marker-free phase maps. Our method's model-free and high-performance nature, combined with its low demand for computational resources, paves the way for accelerating the transition of optical computing from laboratory demonstrations to practical, real-world applications.