Building HMM based unit-selection speech synthesis system using synthetic speech naturalness evaluation score

Heng Lu,Zhen-Hua Ling,Li-Rong Dai,Ren-Hua Wang
DOI: https://doi.org/10.1109/ICASSP.2011.5947567
2011-01-01
ICASSP
Abstract:This paper proposes a unit-selection and waveform concatenation speech synthesis system based on synthetic speech naturalness evaluation. A Support Vector Machine (SVM) and Log Likelihood Ratio (LLR) based synthetic speech naturalness evaluation system was introduced in our previous work. In this paper, the evaluation system is improved in three aspects. Finally, a unit-selection and concatenation waveform speech synthesis system is built on the base of the synthetic speech naturalness evaluation system. Optimum unit sequence is chosen through the re-scoring for the N-best path. Subjective listening tests show the proposed synthetic speech evaluation based speech synthesis system significantly outperforms the traditional unit-selection speech synthesis system.
What problem does this paper attempt to address?