A Differentiable Relaxation of Graph Segmentation and Alignment for AMR Parsing.
Chunchuan LyuShay B. CohenIvan TitovPublished in: CoRR (2020)
Keyphrases
- image segmentation
- graph grammars
- image parsing
- probabilistic relaxation
- segmentation algorithm
- min cut max flow
- objective function
- random walker
- graph model
- graph theory
- random walk
- medical images
- segmentation method
- directed graph
- graph representation
- multiscale
- level set
- loss function
- natural language
- image analysis
- natural language processing
- word alignment
- edge detection
- region merging
- segmentation accuracy
- image alignment
- graph partitioning
- bipartite graph
- region growing
- fully automatic
- graph theoretic
- normalized cut
- fully unsupervised
- graph structure
- segmented regions
- graph matching
- chinese word segmentation
- multi label