Automatic phonetic segmentation by using a SPM-based approach for a Mandarin singing voice corpus.
Cheng-Yuan LinJyh-Shing Roger JangPublished in: INTERSPEECH (2006)
Keyphrases
- fully automatic
- speech recognition
- level set
- image segmentation
- spoken document retrieval
- segmentation algorithm
- semi automatic
- statistical classification
- features extraction
- segmentation method
- topic segmentation
- multiscale
- shape prior
- emotion recognition
- spontaneous speech
- medical images
- hidden markov models
- information retrieval
- prosodic features
- speaker independent
- speech synthesis
- hand crafted
- segmentation accuracy
- graph cuts
- image analysis