Weighted entropy training for the decision tree based text-to-phoneme mapping.
Jilei TianJanne SuontaustaJuha HäkkinenPublished in: INTERSPEECH (2003)
Keyphrases
- training set
- information retrieval
- free text
- speech recognition
- database
- web documents
- minimum error
- training algorithm
- training process
- information theory
- online learning
- keywords
- training samples
- mutual information
- text mining
- text retrieval
- hidden markov models
- speech synthesis
- learning algorithm
- information entropy
- machine learning
- text to speech
- automatically extracted
- training phase
- weighted graph
- pattern recognition
- training examples