Segmentation of continuous speech by HMM and bayesian probability.
Seiichi NakagawaYasuhide HashimotoPublished in: Systems and Computers in Japan (1990)
Keyphrases
- hidden markov models
- minimum description length
- posterior probability
- image segmentation
- segmentation method
- segmentation algorithm
- medical images
- level set
- probability distribution
- multiscale
- image analysis
- energy function
- bayesian networks
- prior probabilities
- bayesian estimation
- shape prior
- posterior distribution
- optimal segmentation
- segmentation errors
- segmentation accuracy
- mr images
- segmented images
- object segmentation
- maximum likelihood
- edge detection
- background subtraction
- medical imaging
- image processing
- speech recognition
- visual speech recognition