Moser Flow: Divergence-based Generative Modeling on Manifolds
arXiv:2108.08052
Abstract
We are interested in learning generative models for complex geometries described via manifolds, such as spheres, tori, and other implicit surfaces. Current extensions of existing (Euclidean) generative models are restricted to specific geometries and typically suffer from high computational costs. We introduce Moser Flow (MF), a new class of generative models within the family of continuous normalizing flows (CNF). MF also produces a CNF via a solution to the change-of-variable formula, however differently from other CNF methods, its model (learned) density is parameterized as the source (prior) density minus the divergence of a neural network (NN). The divergence is a local, linear differential operator, easy to approximate and calculate on manifolds. Therefore, unlike other CNFs, MF does not require invoking or backpropagating through an ODE solver during training. Furthermore, representing the model density explicitly as the divergence of a NN rather than as a solution of an ODE facilitates learning high fidelity densities. Theoretically, we prove that MF constitutes a universal density approximator under suitable assumptions. Empirically, we demonstrate for the first time the use of flow models for sampling from general curved surfaces and achieve significant improvements in density estimation, sample quality, and training complexity over existing CNFs on challenging synthetic geometries and real-world benchmarks from the earth and climate sciences.
References in corpus (13)
- Adam: A Method for Stochastic Optimization
- Language Models are Few-Shot Learners
- Diffusion Models Beat GANs on Image Synthesis
- Variational Inference with Normalizing Flows
- Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains
- Density estimation using Real NVP
- Neural Ordinary Differential Equations
- Normalizing Flows for Probabilistic Modeling and Inference
- Riemannian Continuous Normalizing Flows
- Implicit Geometric Regularization for Learning Shapes
- Neural Manifold Ordinary Differential Equations
- Neural Ordinary Differential Equations on Manifolds
- Latent Variable Modelling with Hyperbolic Normalizing Flows