Unsupervised Font Clustering Using Stochastic Versio of the EM Algorithm and Global Texture Analysis.
Carlos Avilés-CruzJuan Villegas-CortézRené Arechiga-MartínezRafael Escarela-PerezPublished in: CIARP (2004)
Keyphrases
- texture analysis
- em algorithm
- expectation maximization
- unsupervised learning
- finite mixture model
- mixture model
- model based clustering
- minimum message length
- font recognition
- mixture modeling
- finite mixture models
- maximum likelihood
- texture classification
- k means
- multiresolution
- matrix factorisation
- image analysis
- parameter estimation
- gray level
- texture features
- maximum likelihood estimation
- gaussian mixture model
- co occurrence
- density estimation
- texture discrimination
- clustering algorithm
- generative model
- maximum a posteriori
- gabor filters
- expectation maximisation
- gaussian mixture
- pattern recognition
- clustering method
- feature extraction
- image segmentation
- probability density function
- data clustering
- semi supervised
- data points
- supervised learning
- outlier detection
- information theoretic
- probabilistic model
- pairwise
- document clustering
- decision trees