Timestamp-Supervised Action Segmentation with Graph Convolutional Networks.
Hamza KhanSanjay HareshAwais AhmedShakeeb SiddiquiAndrey KoninM. Zeeshan ZiaQuoc-Huy TranPublished in: CoRR (2022)
Keyphrases
- average degree
- image segmentation
- segmentation algorithm
- min cut max flow
- graph structures
- level set
- overlapping communities
- fully connected
- edge weights
- graph structure
- normalized cut
- graph representation
- graph model
- graph theory
- shape prior
- segmentation method
- medical images
- graph layout
- learning algorithm
- weighted graph
- min cut
- complex networks
- energy function
- fully unsupervised
- social networks
- directed edges
- multiscale
- community discovery
- semi supervised
- small world
- unsupervised learning
- segmented regions
- network analysis
- pairwise
- connected components
- random walker
- directed graph
- region growing
- graph matching