Deep Neural Networks motivated by Partial Differential Equations.
Lars RuthottoEldad HaberPublished in: CoRR (2018)
Keyphrases
- partial differential equations
- neural network
- anisotropic diffusion
- level set
- numerical solution
- image denoising
- image processing
- image enhancement
- pattern recognition
- multiscale
- numerical algorithms
- energy functional
- numerical methods
- fourth order
- mathematical morphology
- high order
- numerical scheme
- diffusion equation
- inverse problems
- boundary value problem
- differential equations
- problems in image processing
- conservation laws
- difference equations
- finite difference
- denoising
- natural images
- wavelet transform
- finite difference method
- lattice boltzmann
- nonlinear partial differential equations