Fully automatic segmentation for prosodic speech corpora.
Sarah HoffmannBeat PfisterPublished in: INTERSPEECH (2010)
Keyphrases
- fully automatic segmentation
- speech recognition
- text to speech synthesis
- automatic segmentation
- prosodic features
- text to speech
- manual segmentation
- speech synthesis
- segmentation method
- interactive image segmentation
- synthesized speech
- natural language processing
- speech signal
- speaker verification
- spontaneous speech
- fully automated
- ground truth
- segmentation algorithm
- language model
- contrast enhanced
- gold standard
- ct data
- magnetic resonance images
- human experts
- mr images
- graph cuts
- high quality