Improving the robustness of phonetic segmentation to accent and style variation with a two-staged approach.
Vaishali PatilShrikant JoshiPreeti RaoPublished in: INTERSPEECH (2009)
Keyphrases
- speech recognition
- level set
- fully automatic
- image segmentation
- segmentation algorithm
- spoken language
- hidden markov models
- segmentation method
- segmentation errors
- segmentation accuracy
- energy function
- computational efficiency
- language model
- shape prior
- medical images
- shape variations
- pattern recognition
- data sets
- image regions
- test images
- segmented images
- fully unsupervised
- information retrieval