Bridging the Gap Between Spectral and Spatial Domains in Graph Neural Networks.
Muhammet BalcilarGuillaume RentonPierre HérouxBenoit GaüzèreSébastien AdamPaul HoneinePublished in: CoRR (2020)
Keyphrases
- neural network
- spatial frequency
- pattern recognition
- normalized cut
- spectral characteristics
- spatial information
- spectral decomposition
- artificial neural networks
- spatio temporal
- directed graph
- fuzzy logic
- spatial and temporal
- real world
- graph representation
- application domains
- random walk
- spectral methods
- graph matching
- adjacency matrix
- spatial relations
- graph theory
- graph structure
- graph model
- genetic algorithm
- graph construction
- graph theoretic
- multispectral images
- fault diagnosis
- directed acyclic graph
- multilayer perceptron
- neural network model
- spatial distribution
- graph partitioning
- weighted graph
- bipartite graph
- space time
- spectral analysis
- graph clustering
- neural nets
- spatial relationships
- causal graph
- laplacian matrix
- image segmentation
- multispectral