Hybrid Quantum–Classical Generative Adversarial Network for High-Resolution Image Generation

TLDR

Quantum machine learning promises to outperform classical methods, and quantum generative adversarial networks (QGANs) are a quantum analogue of classical GANs, but existing QGANs are limited to small, down‑scaled images. The authors propose a hybrid quantum‑classical GAN framework that integrates classical and quantum techniques. The framework employs a quantum generator whose qubit count, patch size, layer depth, patch shape, and prior distribution are systematically varied to assess performance. The hybrid GAN generates 28×28 MNIST and Fashion MNIST images with performance comparable to classical GANs while using three orders of magnitude fewer trainable parameters, and larger quantum generators further improve learning.

Abstract

Quantum machine learning (QML) has received increasing attention due to its potential to outperform classical machine learning methods in problems such as classification and identification tasks. A subclass of QML methods is quantum generative adversarial networks (QGANs) which have been studied as a quantum counterpart of classical GANs widely used in image manipulation and generation tasks. The existing work on QGANs is still limited to small-scale proof-of-concept examples based on images with significant down-scaling. Here we integrate classical and quantum techniques to propose a new hybrid quantum-classical GAN framework. We demonstrate its superior learning capabilities over existing quantum techniques by generating 28×28 pixels grey-scale images without dimensionality reduction or classical pre/post-processing on multiple classes of the standard MNIST and Fashion MNIST datasets, which achieves comparable results to classical frameworks with three orders of magnitude less trainable generator parameters. To gain further insight into the working of our hybrid approach, we systematically explore the impact of its parameter space by varying the number of qubits, the size of image patches, the number of layers in the generator, the shape of the patches and the choice of prior distribution. Our results show that increasing the quantum generator size generally improves the learning capability of the network. The developed framework provides a foundation for future design of QGANs with optimal parameter set tailored for complex image generation tasks