Diffusion bridges for stochastic Hamiltonian systems and shape evolutions
arXiv:2002.00885 · doi:10.1137/21M1406283
Abstract
Stochastically evolving geometric systems are studied in shape analysis and computational anatomy for modelling random evolutions of human organ shapes. The notion of geodesic paths between shapes is central to shape analysis and has a natural generalisation as diffusion bridges in a stochastic setting. Simulation of such bridges is key to solve inference and registration problems in shape analysis. We demonstrate how to apply state-of-the-art diffusion bridge simulation methods to recently introduced stochastic shape deformation models thereby substantially expanding the applicability of such models. We exemplify these methods by estimating template shapes from observed shape configurations while simultaneously learning model parameters.
References in corpus (13)
- An integrable shallow water equation with peaked solitons
- Variational Principles for Stochastic Fluid Dynamics
- Overview of the Geometries of Shape Spaces and Diffeomorphism Groups
- Forward-Mode Automatic Differentiation in Julia
- Soliton Dynamics in Computational Anatomy
- Guided proposals for simulating multi-dimensional diffusion bridges
- Stochastic Discrete Hamiltonian Variational Integrators
- Bayesian estimation of discretely observed multi-dimensional diffusion processes using guided proposals
- A Geometric Framework for Stochastic Shape Analysis
- Continuous-discrete smoothing of diffusions
- Simulation of elliptic and hypo-elliptic conditional diffusions
- Conditioning diffusions with respect to partial observations
- Hamiltonian Monte Carlo on Symmetric and Homogeneous Spaces via Symplectic Reduction