Multi-scale Physical Representations for Approximating PDE Solutions with Graph Neural Operators.
Léon MigusYuan YinJocelyn Ahmed MazariPatrick GallinariPublished in: CoRR (2022)
Keyphrases
- multiscale
- partial differential equations
- neural network
- structured data
- real world
- network architecture
- image representation
- graph representations
- candidate solutions
- scale spaces
- graph representation
- graph partitioning
- morphological operators
- directed acyclic graph
- anisotropic diffusion
- graph structure
- bipartite graph
- coarse to fine
- random walk
- natural images
- scale space
- image processing
- connected components
- level set
- optic flow
- graph theoretic
- edge detection
- object recognition
- differential operators