An Hybrid Quantum-Classical Diffusion Model for Image Generation
For researchers in quantum machine learning, this work addresses the qubit cost bottleneck of quantum diffusion models by introducing a hybrid pipeline, though the demonstration is limited to a simple dataset and the results are preliminary.
This paper proposes a hybrid quantum-classical diffusion model that uses a classical autoencoder for dimensionality reduction and a mixed-state quantum denoising diffusion probabilistic model in the latent space, enabling image generation with limited qubit budgets. The method is demonstrated on MNIST, achieving generative capability under realistic hardware constraints.
Quantum diffusion models provide a physics-consistent route to generative learning by formulating noising and denoising directly on quantum states. However, applying such models to classical high-dimensional data is constrained by the qubit cost of state encoding and the computational burden of simulating large density operators. We propose a scalable hybrid generative pipeline that combines a classical autoencoder for dimensionality reduction with a mixed-state quantum denoising diffusion probabilistic model (MSQuDDPM) operating in the learned latent space. The autoencoder compresses data into compact latent codes that can be embedded into a small-qubit Hilbert space, after which the quantum diffusion model learns a generative distribution over latent density operators and decodes samples back to the original domain. Algorithmically, we simplify the reverse dynamics by predicting an estimate of the clean state $ρ_0$ at timestep $t$ and computing the one-step reverse update via an analytic backward propagation rule, rather than learning an explicit predictor for $ρ_{t-1}$. We demonstrate the proposed approach on MNIST image generation and discuss how mixed-state quantum diffusion can serve as a practical backbone for hybrid quantum--classical generative modeling under realistic qubit budgets.