Stochastic Normalizing Flows for Inverse Problems: A Markov Chains Viewpoint.
Paul HagemannJohannes HertrichGabriele SteidlPublished in: SIAM/ASA J. Uncertain. Quantification (2022)
Keyphrases
- markov chain
- inverse problems
- viewpoint
- stochastic process
- probabilistic automata
- markov processes
- monte carlo
- image reconstruction
- global optimization
- convex optimization
- steady state
- transition probabilities
- markov process
- finite state
- state space
- optimization methods
- optimization problems
- early vision
- stationary distribution
- finite automata
- random walk
- partial differential equations
- smoothness constraint
- transition matrix
- stochastic model
- optimization algorithm
- multi objective
- relative entropy
- image analysis