HMM-based unit selection speech synthesis using log likelihood ratios derived from perceptual data
Keywords: Speech synthesis; Unit selection; Hidden Markov model; Log likelihood ratio; Perceptual data