Training of photonic neural networks through in situ backpropagation

TL;DR

This paper presents a novel in situ training method for photonic neural networks using an adapted backpropagation algorithm derived from adjoint variable methods, enabling efficient gradient computation through intensity measurements.

Contribution

It introduces the first efficient in situ training protocol for photonic neural networks, leveraging a photonic analogue of backpropagation for gradient calculation.

Findings

Demonstrated training of a simulated photonic neural network.

Derived exact gradients via intensity measurements within the device.

Potential applications in photonic system optimization and sensitivity analysis.

Abstract

Recently, integrated optics has gained interest as a hardware platform for implementing machine learning algorithms. Of particular interest are artificial neural networks, since matrix-vector multi- plications, which are used heavily in artificial neural networks, can be done efficiently in photonic circuits. The training of an artificial neural network is a crucial step in its application. However, currently on the integrated photonics platform there is no efficient protocol for the training of these networks. In this work, we introduce a method that enables highly efficient, in situ training of a photonic neural network. We use adjoint variable methods to derive the photonic analogue of the backpropagation algorithm, which is the standard method for computing gradients of conventional neural networks. We further show how these gradients may be obtained exactly by performing intensity measurements within the device. As an application, we demonstrate the training of a numerically simulated photonic artificial neural network. Beyond the training of photonic machine learning implementations, our method may also be of broad interest to experimental sensitivity analysis of photonic systems and the optimization of reconfigurable optics platforms.