MAGE 2.0: New Features and its Application in the Development of a Talking Guitar

Maria Astrinaki, Nicolas d'Alessandro, Loïc Reboursière, Alexis Moinet, and Thierry Dutoit

Proceedings of the International Conference on New Interfaces for Musical Expression

Abstract:

This paper describes the recent progress in our approach to generateperformative and controllable speech. The goal of the performative HMM-basedspeech and singing synthesis library, called Mage, is to have the ability togenerate natural sounding speech with arbitrary speaker's voicecharacteristics, speaking styles and expressions and at the same time to haveaccurate reactive user control over all the available production levels. Mageallows to arbitrarily change between voices, control speaking style or vocalidentity, manipulate voice characteristics or alter the targeted contexton-the-fly and also maintain the naturalness and intelligibility of the output.To achieve these controls, it was essential to redesign and improve the initiallibrary. This paper focuses on the improvements of the architectural design,the additional user controls and provides an overview of a prototype, where aguitar is used to reactively control the generation of a synthetic voice invarious levels.

Citation:

Maria Astrinaki, Nicolas d'Alessandro, Loïc Reboursière, Alexis Moinet, and Thierry Dutoit. 2013. MAGE 2.0: New Features and its Application in the Development of a Talking Guitar. Proceedings of the International Conference on New Interfaces for Musical Expression. DOI: 10.5281/zenodo.1178467

BibTeX Entry:

  @inproceedings{Astrinaki2013,
 abstract = {This paper describes the recent progress in our approach to generateperformative and controllable speech. The goal of the performative HMM-basedspeech and singing synthesis library, called Mage, is to have the ability togenerate natural sounding speech with arbitrary speaker's voicecharacteristics, speaking styles and expressions and at the same time to haveaccurate reactive user control over all the available production levels. Mageallows to arbitrarily change between voices, control speaking style or vocalidentity, manipulate voice characteristics or alter the targeted contexton-the-fly and also maintain the naturalness and intelligibility of the output.To achieve these controls, it was essential to redesign and improve the initiallibrary. This paper focuses on the improvements of the architectural design,the additional user controls and provides an overview of a prototype, where aguitar is used to reactively control the generation of a synthetic voice invarious levels.},
 address = {Daejeon, Republic of Korea},
 author = {Maria Astrinaki and Nicolas d'Alessandro and Lo{\"i}c Reboursi{\`e}re and Alexis Moinet and Thierry Dutoit},
 booktitle = {Proceedings of the International Conference on New Interfaces for Musical Expression},
 doi = {10.5281/zenodo.1178467},
 issn = {2220-4806},
 keywords = {speech synthesis, augmented guitar, hexaphonic guitar},
 month = {May},
 pages = {547--550},
 publisher = {Graduate School of Culture Technology, KAIST},
 title = {MAGE 2.0: New Features and its Application in the Development of a Talking Guitar},
 url = {http://www.nime.org/proceedings/2013/nime2013_214.pdf},
 year = {2013}
}