Abstract:Generative models realized with machine learning techniques are powerful tools to infer complex and unknown data distributions from a finite number of training samples in order to produce new synthetic data. Diffusion models are an emerging framework that have recently overcome the performance of the generative adversarial networks in creating synthetic text and high-quality images. Here, we propose and discuss the quantum generalization of diffusion models, i.e., three quantum-noise-driven generative diffusion models that could be experimentally tested on real quantum systems. The idea is to harness unique quantum features, in particular the non-trivial interplay among coherence, entanglement and noise that the currently available noisy quantum processors do unavoidably suffer from, in order to overcome the main computational burdens of classical diffusion models during inference. Hence, we suggest to exploit quantum noise not as an issue to be detected and solved but instead as a very remarkably beneficial key ingredient to generate much more complex probability distributions that would be difficult or even impossible to express classically, and from which a quantum processor might sample more efficiently than a classical one. An example of numerical simulations for an hybrid classical-quantum generative diffusion model is also included. Therefore, our results are expected to pave the way for new quantum-inspired or quantum-based generative diffusion algorithms addressing more powerfully classical tasks as data generation/prediction with widespread real-world applications ranging from climate forecasting to neuroscience, from traffic flow analysis to financial forecasting.

Diffusion models learn distributions generated by complex Langevin dynamics

Lecture Notes in Probabilistic Diffusion Models

An Overview of Diffusion Models: Applications, Guided Generation, Statistical Rates and Optimization

Nonequilbrium physics of generative diffusion models

Where to Diffuse, How to Diffuse, and How to Get Back: Automated Learning for Multivariate Diffusions

Generative Diffusion Models for Lattice Field Theory

Physics-Informed Diffusion Models

Renormalizing Diffusion Models

On the Generalization Properties of Diffusion Models

Unraveling the Smoothness Properties of Diffusion Models: A Gaussian Mixture Perspective

Diffusion Models: A Comprehensive Survey of Methods and Applications

Latent Schr{ö}dinger Bridge Diffusion Model for Generative Learning

Towards Controllable Diffusion Models via Reward-Guided Exploration

Quantum-Noise-Driven Generative Diffusion Models

Discrete generative diffusion models without stochastic differential equations: a tensor network approach

New algorithms for sampling and diffusion models

A Survey on Generative Diffusion Models

Diffusion Probabilistic Fields

Data augmentation-based statistical inference of diffusion processes

A Flexible Diffusion Model