PK-GCN: Prior Knowledge Assisted Image Classification using Graph Convolution Networks.
Xueli XiaoChunyan JiThosini Bamunu MudiyanselageYi PanPublished in: CoRR (2020)
Keyphrases
- image classification
- prior knowledge
- average degree
- graph representation
- dynamic networks
- bag of words
- image representation
- random walk
- overlapping communities
- network size
- small world
- weighted graph
- edge weights
- feature extraction
- highly connected
- graph structures
- graph matching
- graph theory
- graph structure
- community discovery
- graph model
- graph theoretic
- image processing
- directed edges
- multiscale
- structured data
- degree distribution
- network structure
- path length
- community detection
- network analysis
- directed acyclic graph
- clustering coefficient
- graph mining
- link formation
- training data
- graph databases
- social graphs
- real world graphs
- spanning tree
- complex networks
- connected components
- directed graph
- scale free
- class specific
- prior information