Riemannian Denoising Diffusion Probabilistic Models

Abstract

We propose Riemannian Denoising Diffusion Probabilistic Models (RDDPMs) for learning distributions on submanifolds of Euclidean space that are level sets of functions, including most of the manifolds relevant to applications. Existing methods for generative modeling on manifolds rely on substantial geometric information such as geodesic curves or eigenfunctions of the Laplace-Beltrami operator and, as a result, they are limited to manifolds where such information is available. In contrast, our method, built on a projection scheme, can be applied to more general manifolds, as it only requires being able to evaluate the value and the first order derivatives of the function that defines the submanifold. We provide a theoretical analysis of our method in the continuous-time limit, which elucidates the connection between our RDDPMs and score-based generative models on manifolds. The capability of our method is demonstrated on datasets from previous studies and on new datasets sampled from two high-dimensional manifolds, i.e. $\mathrm{SO}(10)$ and the configuration space of molecular system alanine dipeptide with fixed dihedral angle.

References in corpus (13)

• Denoising Diffusion Probabilistic Models
• Deep Unsupervised Learning using Nonequilibrium Thermodynamics
• Generative Modeling by Estimating Gradients of the Data Distribution
• GeoDiff: a Geometric Diffusion Model for Molecular Conformation Generation
• Implicit Geometric Regularization for Learning Shapes
• Riemannian Continuous Normalizing Flows
• Neural Manifold Ordinary Differential Equations
• Riemannian Score-Based Generative Modelling
• Riemannian Diffusion Models
• Moser Flow: Divergence-based Generative Modeling on Manifolds
• Scaling Riemannian Diffusion Models
• Ergodic SDEs on submanifolds and related numerical sampling schemes
• Generative Modeling on Manifolds Through Mixture of Riemannian Diffusion Processes