Training mixture of weighted SVM for object detection using EM algorithm.
De ChengJinjun WangXing WeiYihong GongPublished in: Neurocomputing (2015)
Keyphrases
- em algorithm
- expectation maximization
- mixture model
- object detection
- training procedure
- object detectors
- gaussian mixture model
- gaussian mixture
- mixture modeling
- finite mixture model
- parameter estimation
- maximum likelihood
- training set
- mixture distribution
- estimate the model parameters
- support vector
- finite mixture models
- multi class
- maximum likelihood estimation
- probabilistic model
- generative model
- incomplete data
- mixture components
- support vector machine
- feature selection
- likelihood function
- probability density function
- training examples
- maximum a posteriori
- expectation maximisation
- penalized likelihood
- density estimation
- object recognition
- knn
- hyperparameters
- supervised learning
- computer vision
- object categories
- gaussian distribution
- maximum likelihood estimates
- machine learning
- training samples
- image segmentation
- training data
- feature vectors
- structured prediction
- log likelihood
- k means
- mixture of gaussians
- model selection
- feature space
- selection criterion
- image processing
- maximum margin
- unsupervised learning