Photonic quantum generative adversarial networks for classical data

TL;DR

This paper introduces a photonic quantum GAN using linear optical circuits and Fock-space encoding, capable of generating images, and demonstrates its training on a single-photon quantum processor, advancing near-term quantum generative models.

Contribution

It presents a novel photonic quantum GAN architecture compatible with near-term quantum hardware, demonstrating end-to-end training on a single-photon processor.

Findings

Successfully trained the model to generate images.

Demonstrated compatibility with near-term photonic quantum computing.

Achieved end-to-end training on a single-photon quantum processor.

Abstract

In generative learning, models are trained to produce new samples that follow the distribution of the target data. These models were historically difficult to train, until proposals such as Generative Adversarial Networks (GANs) emerged, where a generative and a discriminative model compete against each other in a minimax game. Quantum versions of the algorithm were since designed, both for the generation of classical and quantum data. While most work so far has focused on qubit-based architectures, in this article we present a quantum GAN based on linear optical circuits and Fock-space encoding, which makes it compatible with near-term photonic quantum computing. We demonstrate that the model can learn to generate images by training the model end-to-end experimentally on a single-photon quantum processor.