Applying the extended mass-constraint EM algorithm to image retrieval.
Daan HeNick CerconeZhenmei GuPublished in: Comput. Math. Appl. (2008)
Keyphrases
- em algorithm
- image retrieval
- expectation maximization
- mixture model
- maximum likelihood
- generative model
- parameter estimation
- maximum likelihood estimation
- gaussian mixture model
- gaussian mixture
- expectation maximisation
- mixture modeling
- maximum a posteriori
- likelihood function
- probability density function
- incomplete data
- log likelihood
- image content
- low level features
- relevance feedback
- probabilistic model
- object recognition
- penalized likelihood
- log likelihood function
- mixture distribution
- probabilistic principal component analysis
- finite mixtures
- finite mixture models
- machine learning
- likelihood maximization
- finite mixture model
- model based clustering
- visual features
- markov random field
- feature selection
- information retrieval