Unsupervised Prosodic Labeling of Speech Synthesis Databases Using Context-Dependent HMMs.
Chen-Yu YangZhen-Hua LingLi-Rong DaiPublished in: IEICE Trans. Inf. Syst. (2014)
Keyphrases
- context dependent
- speech synthesis
- speech recognition
- databases
- prosodic features
- hidden markov models
- text to speech
- semantic level
- unsupervised learning
- vocal tract
- context free
- low level
- speech corpus
- speech signal
- pattern recognition
- active learning
- supervised learning
- database
- knowledge discovery
- language model
- semi supervised
- user centric
- natural language
- database systems
- automatic speech recognition
- noisy environments
- relational databases
- image segmentation
- information extraction
- sequential data
- neural network
- data model