An unsupervised audio segmentation method using Bayesian information criterion.
Ezgi Can OzanSeda TankizBanu Oskay AcarTolga ÇilogluPublished in: ISCCSP (2014)
Keyphrases
- segmentation method
- bayesian information criterion
- model selection
- gaussian mixture model
- information criterion
- mixture model
- unsupervised learning
- region growing
- object segmentation
- image segmentation
- akaike information criterion
- speaker diarization
- active contours
- selection criterion
- segmentation algorithm
- automated segmentation
- model based segmentation
- segmented images
- active contour model
- shape prior
- supervised learning
- energy function
- segmentation result
- watershed transform
- superpixels
- conditional random fields
- region merging
- ultrasound image segmentation
- semi supervised
- intensity information
- machine learning
- learning algorithm
- multiscale
- probabilistic model
- segmentation procedure
- expectation maximization
- broadcast news