Analyzing partially-polarized light with a photonic deep random neural network

Abstract

Optical neural networks are emerging as a powerful and versatile tool for processing optical signals directly in the optical domain with superior speed, integrability, and functionality. Their application to optical polarization enables neuromorphic polarization sensors, but their operation is limited to fully-polarized light. Here, we demonstrate single-shot analysis of partially-polarized beams with a photonic random neural network (PRNN). The PRNN is composed of a series of optical layers implemented by a stack of scattering media and a few trainable digital nodes. The setup infers the degree-of-polarization and the Stokes parameters of the polarized component with precision comparable to off-the-shelf polarimeters. The use of several optical layers allows to enhance the accuracy, reduce the sensor size, and minimize digital costs, demonstrating the advantage of a deep optical encoder for processing polarization information. Our results point out photonic neural networks as fast, compact, broadband, low-cost polarimeters that are widely applicable from sensing to imaging.