Maximum-likelihood dimensionality reduction in gaussian mixture models with an application to object classification.
Massimo PiccardiHatice GunesAhmed Fawzi OtoomPublished in: ICPR (2008)
Keyphrases
- object classification
- gaussian mixture model
- maximum likelihood
- dimensionality reduction
- em algorithm
- mixture model
- expectation maximization
- feature extraction
- high dimensional
- high dimensional data
- object recognition
- object detection
- gaussian distribution
- low dimensional
- principal component analysis
- object categories
- data points
- pedestrian detection
- unsupervised learning
- feature space
- log likelihood function
- variational bayes
- pattern recognition
- benchmark datasets
- face recognition
- speaker identification
- sparse representation
- visual words
- speaker recognition
- feature selection
- machine learning
- generative model
- language model
- feature set
- image segmentation