UGrAD: A graph-theoretic framework for classification of activity with complementary graph boundary detection.
Tamal BatabyalScott T. ActonAndrea VaccariPublished in: ICIP (2016)
Keyphrases
- object recognition
- boundary detection
- theoretic framework
- graph matching
- object detection and recognition
- computer vision
- spectral graph
- graph representation
- directed graph
- structured data
- information theoretic
- graph mining
- pattern recognition
- support vector
- graph databases
- weighted graph
- directed acyclic graph
- feature extraction
- graph structure
- decision trees
- three dimensional
- graph model
- mutual information
- graph theory
- bipartite graph
- class labels
- support vector machine svm
- text classification