Combining Gaussian Mixture Models and Support Vector Machines for Relevance Feedback in Content Based Image Retrieval.
Apostolos MarakakisNikolas P. GalatsanosAristidis LikasAndreas StafylopatisPublished in: AIAI (2009)
Keyphrases
- gaussian mixture model
- relevance feedback
- support vector
- image retrieval
- mixture model
- improve the retrieval accuracy
- learning machines
- large margin classifiers
- information retrieval systems
- image search
- em algorithm
- speaker recognition
- feature vectors
- feature space
- speaker identification
- low level features
- information retrieval
- probability density function
- cbir systems
- expectation maximization
- maximum likelihood
- density estimation
- gaussian mixture
- low level image features
- covariance matrices
- retrieval accuracy
- classification accuracy
- cross validation
- variational bayes
- image database
- multimedia
- learning algorithm
- feature extraction
- probability density
- image classification
- similarity measure
- feature selection