Optimizing a DIscrete Loss (ODIL) to solve forward and inverse problems for partial differential equations using machine learning tools.
Petr KarnakovSergey LitvinovPetros KoumoutsakosPublished in: CoRR (2022)
Keyphrases
- inverse problems
- learning tools
- partial differential equations
- image denoising
- image reconstruction
- global optimization
- anisotropic diffusion
- level set
- image processing
- learning experience
- numerical solution
- convex optimization
- distance learning
- mobile learning
- energy functional
- image enhancement
- optimization methods
- high order
- optimization problems
- early vision
- computer vision
- multiscale
- simulated annealing
- machine learning
- noisy images
- parameter space
- neural network
- image restoration
- denoising
- object recognition