A comparison of open-source segmentation architectures for dealing with imperfect data from the media in speech synthesis.
Ascensión Gallardo-AntolínJuan Manuel MonteroSimon KingPublished in: INTERSPEECH (2014)
Keyphrases
- speech synthesis
- open source
- speech recognition
- segmentation method
- level set
- text to speech
- vocal tract
- multimedia
- segmentation algorithm
- shape prior
- image segmentation
- region growing
- energy function
- prosodic features
- segmentation errors
- fully automatic
- segmentation accuracy
- object segmentation
- grey level
- cross media
- brain mri
- information retrieval
- multimedia data
- hidden markov models
- quantitative evaluation
- word segmentation
- language model
- source code
- markov random field
- image analysis
- computer vision
- data mining
- speech corpus