Content-based music retrieval with nonlinear feature space transformation using relevance feedback.
Shunsuke SakaiKeisuke KameyamaPublished in: SMC (2008)
Keyphrases
- feature space
- relevance feedback
- image retrieval
- high dimensional feature space
- content based music retrieval
- kernel pca
- generalized discriminant analysis
- high dimension
- feature vectors
- input space
- kernel based nonlinear
- mean shift
- training samples
- query expansion
- information retrieval systems
- high dimensional
- multiple features
- high dimensionality
- document retrieval
- active learning
- principal component analysis
- low level features
- low dimensional
- feedback information
- mercer kernels
- feature selection
- relevant documents
- classification accuracy
- kernel methods
- retrieval process
- dimensionality reduction
- user feedback
- feature weighting
- kernel principal component analysis
- information retrieval
- support vector machine
- retrieval model
- machine learning
- kernel function
- image representation
- retrieval effectiveness
- cbir systems
- dimension reduction
- image content
- visual features
- feature extraction
- image database
- input data
- feature set