q-Gaussian Mixture Models Based on Non-extensive Statistics for Image and Video Semantic Indexing.
Nakamasa InoueKoichi ShinodaPublished in: ACCV (2) (2012)
Keyphrases
- gaussian mixture model
- semantic indexing
- image data
- multiscale
- probability density function
- image content
- mixture model
- image classification
- image features
- low level
- image representation
- video sequences
- image retrieval
- em algorithm
- maximum likelihood
- image regions
- feature vectors
- expectation maximization
- key frames
- image segmentation
- image collections
- visual data
- video streams
- multimedia
- semantic description
- multimedia information retrieval
- semantic concepts
- object recognition
- principal component analysis
- generative model
- computer vision