Factored Maximum Penalized Likelihood Kernel Regression for HMM-Based Style-Adaptive Speech Synthesis.
June Sig SungDoo Hwa HongNam Soo KimPublished in: IEEE J. Sel. Top. Signal Process. (2014)
Keyphrases
- kernel regression
- speech synthesis
- penalized likelihood
- speech recognition
- regression model
- text to speech
- hidden markov models
- nadaraya watson
- maximum likelihood
- linear regression
- em algorithm
- vocal tract
- least squares
- neural network
- speech signal
- pattern recognition
- regression methods
- dimension reduction
- log likelihood
- optimization algorithm
- evolutionary algorithm
- machine learning