Embodiment of Learning in Electro-Optical Signal Processors

TL;DR

This paper demonstrates how backpropagation can be physically implemented in electro-optical delay-coupled systems, significantly improving their performance on complex tasks and enabling autonomous training.

Contribution

It introduces a method to physically realize backpropagation in electro-optical systems, enhancing their ability to learn and adapt autonomously.

Findings

Error rate decreases significantly with backpropagation implementation

Electro-optical systems can embody their own training process

Improved performance on benchmark tasks

Abstract

Delay-coupled electro-optical systems have received much attention for their dynamical properties and their potential use in signal processing. In particular it has recently been demonstrated, using the artificial intelligence algorithm known as reservoir computing, that photonic implementations of such systems solve complex tasks such as speech recognition. Here we show how the backpropagation algorithm can be physically implemented on the same electro-optical delay-coupled architecture used for computation with only minor changes to the original design. We find that, compared when the backpropagation algorithm is not used, the error rate of the resulting computing device, evaluated on three benchmark tasks, decreases considerably. This demonstrates that electro-optical analog computers can embody a large part of their own training process, allowing them to be applied to new, more difficult tasks.