Selection of waveform units for corpus-based Mandarin speech synthesis based on decision trees and prosodic modification costs.
Fu-Chiang ChouChiu-yu TsengLin-Shan LeePublished in: EUROSPEECH (1999)
Keyphrases
- speech synthesis
- prosodic features
- speech recognition
- decision trees
- text to speech
- vocal tract
- language model
- speaker verification
- speech corpus
- predictive accuracy
- cost sensitive learning
- data sets
- training data
- automatic speech recognition
- machine learning
- random forest
- speech signal
- rule induction
- noisy environments
- image formation
- hidden markov models
- pattern recognition
- decision tree algorithm
- data mining methods
- decision rules
- naive bayes