Graph Neural Networks for Aerodynamic Flow Reconstruction from Sparse Sensing.
Gregory DuthéImad AbdallahSarah BarberEleni N. ChatziPublished in: CoRR (2023)
Keyphrases
- neural network
- flow patterns
- compressed sensing
- compressive sensing
- signal reconstruction
- graph theory
- random projections
- sensor networks
- image reconstruction
- weighted graph
- gaussian graphical models
- random walk
- graph representation
- graph structure
- artificial neural networks
- sparse data
- directed acyclic graph
- real time
- high resolution
- fuzzy logic
- graph matching
- high dimensional
- three dimensional
- graph model
- multi layer
- bipartite graph
- neural nets
- multilayer perceptron
- connected components
- graph mining
- feed forward
- back propagation
- graph theoretic
- genetic algorithm
- pattern recognition
- reconstruction process
- discrete tomography
- shape recovery
- signal recovery
- sparsity constraints
- flow field
- feature space
- directed graph
- training process