Action Co-localization in an Untrimmed Video by Graph Neural Networks.
Changbo ZhaiLe WangQilin ZhangZhanning GaoZhenxing NiuNanning ZhengGang HuaPublished in: MMM (1) (2020)
Keyphrases
- neural network
- human actions
- activity detection
- pattern recognition
- video sequences
- video data
- action classification
- graph theory
- action detection
- multimedia
- graph structure
- fuzzy logic
- back propagation
- graph model
- video clips
- genetic algorithm
- real time
- recognizing human actions
- video content
- directed graph
- action recognition
- space time
- video frames
- random walk
- video streams
- bipartite graph
- graph representation
- video analysis
- spatial and temporal
- graph matching
- video surveillance
- self organizing maps
- feed forward
- graph theoretic
- neural network model
- weighted graph
- artificial neural networks
- localization algorithm
- digital video
- image sequences
- graph partitioning
- graph mining
- multi layer
- event detection
- neural nets
- recurrent neural networks
- human activities