HMM-based text to speech system with speaker interpolation Konuşmaci aradeǧerlemeli̇ SMM tabanli meti̇nden konuşma sentezleme si̇stemi̇


Orhan M. C., Demiroǧlu C.

2011 IEEE 19th Signal Processing and Communications Applications Conference, SIU 2011, Antalya, Turkey, 20 - 22 April 2011, pp.781-784, (Full Text) identifier

  • Publication Type: Conference Paper / Full Text
  • Doi Number: 10.1109/siu.2011.5929768
  • City: Antalya
  • Country: Turkey
  • Page Numbers: pp.781-784
  • Akdeniz University Affiliated: No

Abstract

In this paper, we propose a joint multiple relay selection and power optimization technique with low computational complexity to improve the error performance of the cooperative communication networks. Proposed method has a reduced computational load for the multiple relay selection problem compared to the optimum solution which is based on selecting the “best” relays. Moreover, since the relay selection is performed iteratively until a threshold SNR is reached, power optimization is also achieved at every iteration. Hence, we attain an improved error performance using the least possible number of relays, and as such low power consumption.

Hidden Markov Model (HMM) based text-to-speech (TTS) systems offer many advantages compared to the concatenative approach. One of those advantages is the ability to interpolate between different speakers to generate new voices. In this paper, speaker interpolation for HMM-based TTS (HTS) is described and listening test results for the interpolation of English and Turkish voices are presented. Similar to English, we obtained Turkish speech that strongly reflect the interpolation ratio in perceptual similarity. Some insight into the interpolation process is also provided by analysing the spectra of the reference and final voices. © 2011 IEEE.