F&lt;sub&gt;0&lt;/sub&gt; contour parametric modeling using multivariate adaptive regression splines for arabic text-to-speech synthesis

Mnasri, Zied; Boukadida, Fatouma; Ellouze, Noureddine

doi:10.1109/ssd.2011.5981479

Cited by 1 publication

(1 citation statement)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Details on Fujisaki model and extraction of model parameters are available in [7]- [11]. In the paper, Mnasri et al [12] have built a neural network model to predict pitch frequency using Fujisaki parameters to synthesize Arabic language. In the paper, Rao and Yegnanarayana [13] has developed a neural network model to predict F0 of a syllable for Telugu and Bengali languages by feeding the attributes obtained directly from the utterances.…”

Section: Introductionmentioning

confidence: 99%

Performance analysis of different intonation models in Kannada speech synthesis

Chakrasali

Indira

Narasimhaiah³

et al. 2022

IJEECS

View full text Add to dashboard Cite

Text <span lang="EN-US">to speech (TTS) is a system that generates artificial speech from text input. The prosodic models used improve the quality of the synthesized speech especially naturalness and intelligibility. The prosody involves intonation, intonation refers to the variations in the pitch frequency (F0) with respect to time in an utterance. This work mainly concentrates on building feedback neural network model to predict F0 contour in the utterances using Fujisaki intonation model parameters as the input features to the network since the Fujisaki intonation model is data driven and not a rule based one. In this work we have built 4-layer feedback neural network in the festival framework. Finally, the synthetically generated Kannada speech using the neural network model, is compared for its performance with the classification and regression tree (CART) model and Tilt model. Database of simple declarative Kannada sentences created by Carnegie Mellon University have been deployed in this work. From the study it is very clear that F0 contours can be accurately predicted using CART and neural network models, whereas naturalness and intelligibility is high in CART model rather than neural network model.</span>

show abstract

Section: Introductionmentioning

confidence: 99%