LDA Versus MMD Approximation on Mislabeled Images for Dependant Selection of Visual Features and Their Heterogeneity.
Sabrina TollariHervé GlotinPublished in: ICASSP (2) (2006)
Keyphrases
- visual features
- image collections
- image classification
- image retrieval
- image annotation
- visual appearance
- web images
- visually similar
- visual information
- low level visual features
- semantic gap
- visual content
- visual similarity
- sample images
- labeled images
- visual data
- global features
- cbir systems
- visual patterns
- bridge the semantic gap
- bag of features
- image search
- image database
- image features
- probabilistic latent semantic analysis
- visual descriptors
- low level features
- semantic concepts
- low level
- visual attributes
- visual and textual features
- object recognition
- feature extraction
- automatic image annotation
- keywords
- semantic features
- image set
- image content
- spatial information
- input image
- bag of words
- image representation
- principal component analysis
- feature space
- visual properties
- search engine