Abstract:Score-based stochastic denoising models have recently been demonstrated as powerful machine learning tools for conditional and unconditional image generation. The existing methods are based on a forward stochastic process wherein the training images are scaled to zero over time and white noise is gradually added such that the final time step is approximately zero-mean identity-covariance Gaussian noise. A neural network is then trained to approximate the time-dependent score function, or the gradient of the logarithm of the probability density, for that time step. Using this score estimator, it is possible to run an approximation of the time-reversed stochastic process to sample new images from the training data distribution. These score-based generative models have been shown to out-perform generative adversarial neural networks using standard benchmarks and metrics. However, one issue with this approach is that it requires a large number of forward passes of the neural network. Additionally, the images at intermediate time steps are not useful, since the signal-to-noise ratio is low. In this work we present a new method called Fourier Diffusion Models which replaces the scalar operations of the forward process with shift-invariant convolutions and the additive white noise with additive stationary noise. This allows for control of MTF and NPS at intermediate time steps. Additionally, the forward process can be crafted to converge to the same MTF and NPS as the measured images. This way, we can model continuous probability flow from true images to measurements. In this way, the sample time can be used to control the tradeoffs between measurement uncertainty and generative uncertainty of posterior estimates. We compare Fourier diffusion models to existing scalar diffusion models and show that they achieve a higher level of performance and allow for a smaller number of time steps.

Score-based Diffusion Models for Generating Liquid Argon Time Projection Chamber Images

Score-based diffusion models for generating liquid argon time projection chamber images

Fourier Diffusion Models: A Method to Control MTF and NPS in Score-Based Stochastic Image Generation

Scalable Deep Convolutional Neural Networks for Sparse, Locally Dense Liquid Argon Time Projection Chamber Data

Towards Designing Scalable Quantum-Enhanced Generative Networks for Neutrino Physics Experiments with Liquid Argon Time Projection Chambers

Feature-guided score diffusion for sampling conditional densities

LN3Diff: Scalable Latent Neural Fields Diffusion for Speedy 3D Generation

Score-based Denoising Diffusion with Non-Isotropic Gaussian Noise Models

Fast Point Cloud Generation with Diffusion Models in High Energy Physics

Score-based generative diffusion with "active" correlated noise sources

Scalable, Proposal-free Instance Segmentation Network for 3D Pixel Clustering and Particle Trajectory Reconstruction in Liquid Argon Time Projection Chambers

Score-based Diffusion Models for Bayesian Image Reconstruction

4Diffusion: Multi-view Video Diffusion Model for 4D Generation

Multilevel Diffusion: Infinite Dimensional Score-Based Diffusion Models for Image Generation

Diffusion$^2$: Dynamic 3D Content Generation via Score Composition of Video and Multi-view Diffusion Models

Diffusion model approach to simulating electron-proton scattering events

Point cloud-based diffusion models for the Electron-Ion Collider

L3DG: Latent 3D Gaussian Diffusion

Baking Gaussian Splatting into Diffusion Denoiser for Fast and Scalable Single-stage Image-to-3D Generation

Score-Based Diffusion Models as Principled Priors for Inverse Imaging

Neural Point Cloud Diffusion for Disentangled 3D Shape and Appearance Generation