Multi-label semantic concept detection in videos using fusion of asymmetrically trained deep convolutional neural networks and foreground driven concept co-occurrence matrix.
Nitin J. JanweKishor K. BhoyarPublished in: Appl. Intell. (2018)
Keyphrases
- co occurrence
- multi label
- semantic concept detection
- semantic concepts
- image annotation
- semantic similarity
- convolutional neural networks
- text categorization
- visual features
- image classification
- concept detectors
- semantic gap
- multi modal
- graph cuts
- wordnet
- visual concepts
- text classification
- video analysis
- semantic information
- concept detection
- named entities
- image search
- text documents
- class labels
- low level features
- training set
- video sequences
- image understanding
- image collections
- multimedia content
- object categories
- visual words
- low level
- visual content
- machine learning
- text mining
- multiscale
- high level