BI-GCN: Boundary-Aware Input-Dependent Graph Convolution Network for Biomedical Image Segmentation.
Yanda MengHongrun ZhangDongxu GaoYitian ZhaoXiaoyun YangXuesheng QianXiaowei HuangYalin ZhengPublished in: BMVC (2021)
Keyphrases
- image segmentation
- image processing
- graph partitioning
- boundary and region information
- object boundaries
- random walk
- peer to peer
- graph cuts
- fully connected
- network model
- graph theory
- business intelligence
- link prediction
- deformable models
- information extraction
- computer vision
- active contours
- graphical representation
- graph representation
- spanning tree
- graph model
- graph mining
- segmentation result
- input patterns
- structured data
- boundary information
- connected components
- network structure
- segmentation method
- wireless sensor networks
- strongly connected
- biomedical literature
- normalized cut
- text mining
- boundary detection
- clustering algorithm
- energy functional
- weighted graph
- shortest path
- multiscale
- directed graph
- binary images