Prosody modeling for syllable based text-to-speech synthesis using feedforward neural networks

摘要

Prosody plays an important role in improving the quality of text-to-speech synthesis (TTS) system. In this paper, features related to the linguistic and the production constraints are proposed for modeling the prosodic parameters such as duration, intonation and intensities of the syllables. The linguistic constraints are represented by positional, contextual and phonological features, and the production constraints are represented by articulatory features. Neural network models are explored to capture the implicit duration, F-0 and intensity knowledge using above mentioned features. The prediction performance of the proposed neural network models is evaluated using objective measures such as average prediction error (mu), standard deviation (sigma) and linear correlation coefficient (gamma(X,Y)). The prediction accuracy of the proposed neural network models is compared with other state-of-the-art prosody models used in TTS systems. The prediction accuracy of the proposed prosody models is also verified by conducting listening tests, after integrating the proposed prosody models to the baseline TTS system. (C) 2015 Elsevier B.V. All rights reserved.