Scaling Quantum Machine Learning without Tricks: High-Resolution and Diverse Image Generation

TL;DR

This paper demonstrates a novel quantum generative model capable of producing high-resolution, diverse images across multiple datasets without relying on traditional tricks, marking a significant advancement in quantum machine learning.

Contribution

The authors introduce a quantum Wasserstein GAN that generates full-resolution images from complete datasets, surpassing previous limitations and establishing a new state-of-the-art in quantum image generation.

Findings

Achieved high-resolution image generation on MNIST and Fashion-MNIST datasets.

Extended approach to color images with Street View House Numbers dataset.

Maintained performance under quantum shot noise conditions.

Abstract

Quantum generative modeling is a rapidly evolving discipline at the intersection of quantum computing and machine learning. Contemporary quantum machine learning is generally limited to toy examples or heavily restricted datasets with few elements. This is not only due to the current limitations of available quantum hardware but also due to the absence of inductive biases arising from application-agnostic designs. Current quantum solutions must resort to tricks to scale down high-resolution images, such as relying heavily on dimensionality reduction or utilizing multiple quantum models for low-resolution image patches. Building on recent developments in classical image loading to quantum computers, we circumvent these limitations and train quantum Wasserstein GANs on the established classical MNIST and Fashion-MNIST datasets. Using the complete datasets, our system generates full-resolution images across all ten classes and establishes a new state-of-the-art performance with a single end-to-end quantum generator without tricks. As a proof-of-principle, we also demonstrate that our approach can be extended to color images, exemplified on the Street View House Numbers dataset. We analyze how the choice of variational circuit architecture introduces inductive biases, which crucially unlock this performance. Furthermore, enhanced noise input techniques enable highly diverse image generation while maintaining quality. Finally, we show promising results even under quantum shot noise conditions.