Prediction of sentence importance for speech summarization using prosodic parameters.
Akira InoueTakayoshi MikamiYoichi YamashitaPublished in: INTERSPEECH (2003)
Keyphrases
- speech recognition
- text to speech synthesis
- prediction accuracy
- text to speech
- linear regression model
- automatic summarization
- root mean square error
- linear prediction
- prosodic features
- relative importance
- maximum likelihood
- text summarization
- prediction algorithm
- prediction error
- parameter values
- natural language
- neural network
- speech signal
- spoken language
- speaker verification
- prediction model
- parameter estimation
- machine learning