Segmental Duration Modeling in Turkish

Date

2006-01-01

Author

Ozturk, Ozlem
Çiloğlu, Tolga

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

36
views

0
downloads

Naturalness of synthetic speech highly depends on appropriate modeling of prosodic aspects. Mostly, three prosody components are modeled: segmental duration, pitch contour and intensity. In this study, we present our work on modeling segmental duration in Turkish using machine-learning algorithms, especially Classification and Regression Trees (CART). The models predict phone durations based on attributes such as phone identity, neighboring phone identities, lexical stress, position of syllable in word, part-of-speech (POS) information, word length in number of syllables and position of word in utterance extracted from a speech corpus of approximately 700 sentences. Obtained models predict segment durations better than mean duration approximations (similar to 0.77 Correlation Coefficient, CC, and 20.4 ms Root-Mean Squared Error, RMSE). Attributes phone identity, neighboring phone identities, lexical stress, syllable type, POS, phrase break information, and location of word in the phrase constitute best predictor set for phoneme duration modeling.

Subject Keywords

Segmental duration modeling

URI

https://hdl.handle.net/11511/53785

Collections

Department of Electrical and Electronics Engineering, Conference / Seminar

Suggestions

OpenMETU
Core

Segmental duration modelling in Turkish Ozturk, Ozlem; Çiloğlu, Tolga (2006-01-01) Naturalness of synthetic speech highly depends on appropriate modelling of prosodic aspects. Mostly, three prosody components are modelled: segmental duration, pitch contour and intensity. In this study, we present our work on modelling segmental duration in Turkish using machine-learning algorithms, especially Classification and Regression Trees. The models predict phone durations based on attributes such as current, preceding and following phones' identities, stress, part-of-speech, word length in number ...
Near-Earth space plasma modelling and forecasting Strangeways, Hal J.; Kutiev, Ivan; Cander, Ljiljana R.; Kouris, Stamatis; Gherm, Vadim; Marin, Diego; De La Morena, Benito; Pryse, S. Eleri; Perrone, Loredana; Pietrella, Marco; Stankov, Stanimir; Tomasik, Lukasz; Tulunay, Ersin; Tulunay, Yurdanur; Zernov, Nikolay; Zolesi, Bruno (2009-06-01) In the frame of the European COST 296 project (Mitigation of Ionospheric Effects oil Radio Systems, MIERS) in the Working Package 1.3, new ionospheric models, prediction and forecasting methods and programs as well as ionospheric imaging techniques have been developed. They include (i) topside ionosphere and meso-scale irregularity models, (ii) improved forecasting methods for real time forecasting and for prediction of foF2, M(3000)F2, MUF and TECs, including the use of new techniques such as Neurofuzzy, N...
Completion, pricing and calibration in a Lévy market model Yılmaz, Büşra Zeynep; Hayfavi, Azize; Erol, Işıl; Department of Financial Mathematics (2010) In this thesis, modelling with Lévy processes is considered in three parts. In the first part, the general geometric Lévy market model is examined in detail. As such markets are generally incomplete, it is shown that the market can be completed by enlarging with a series of new artificial assets called “power-jump assets” based on the power-jump processes of the underlying Lévy process. The second part of the thesis presents two different methods for pricing European options: the martingale pricing approach...
Non-parametric and semi-parametric regional modeling of the ionospheric vertical total electron content using ground-based GPS observations Durmaz, Murat; Karslıoğlu, Mahmut Onur; Department of Geodetic and Geographical Information Technologies (2013) Different global and regional models have been proposed for modeling of Vertical Total Electron Content(VTEC). Basis functions such as Spherical Harmonics and tensor products of univariate B-Splines are used for global and regional modeling respectively. The coefficients of the basis functions and Differential Code Biases (DCB) of satellites and receivers are treated as parameters and estimated from geometry-free linear combinations. In this work, a new approach based on non-parametric and semi-parametric m...
A model of angiogenesis by hybrid systems with delay on the piecewise constant part Karasözen, Bülent; Kahraman, Mustafa (2009-09-01) An important approach towards understanding the cancer dynamics is the modeling of angiogenesis process. There have been several attempts to model this process. Among them angiogenesis models with time delays, caused by the physical distance between the tumor and the vessel, are the most realistic ones. Recent studies have suggested that those delays can cause oscillatory behavior in the angiogenesis process. In this work we employed piecewise linear hybrid systems with delay on the piecewise constant part....

Citation Formats

O. Ozturk and T. Çiloğlu, “Segmental Duration Modeling in Turkish,” 2006, Accessed: 00, 2020. [Online]. Available: https://hdl.handle.net/11511/53785.