Development of Feature Set, Classification Implementation and Applications for Vowel Migration/Modification in Sung Filipino (Tagalog) Texts and Perceived Intelligibility

  • Virginia B. Bustos University of the Philipppines
  • Triah Joyce G. De la Cruz University of the Philippines
  • Ramon Maria G. Acoymo University of the Philippines
  • Rowena Cristina L. Guevara University of the Philippines

Abstract

With the emergence of research on real-time visual feedback to supplement vocal pedagogy, the utilization of technology in the world of music is now seen to accelerate skills learning and enhance cognitive development. The researchers of this project aim to further analyze vowel intelligibility and develop software applications intended to be used not only by professional singers but also by individuals who wish to improve their singing capability.

 

Data in the form of sung vowels and song pieces were obtained from 46 singers. A Listening Test was then conducted on these samples to obtain the ground truth for vowel classification based on human perception. Simulation of the human auditory perception of sung Filipino vowels was performed using formant frequencies and Mel-frequency cepstral coefficients as feature vector inputs to a two-stage Discriminant Analysis classifier. The setup resulted in an over-all Training Set accuracy of 89.4% and an over-all Test Set accuracy of 90.9%. The accuracy of the classifier, measured in terms of the correspondence of vowel classifications obtained from the classifier with the results of the Listening Test, reached 92.3%. Using information obtained from the classifier, offline and online/real-time software applications were developed.

 

The main application features include the display of the spectral envelope and spectrogram, pitch and vibrato analysis and direct feedback on the classification of the sung vowel. These features were recommended by singers who were surveyed and were incorporated in the applications to aid singers to adjust formant locations, directly determine listener’s perception of sung vowels, perform modeling effectively and carry out vowel migration.

Author Biographies

Virginia B. Bustos, University of the Philipppines
Electrical and Electronics Engineering Insitute
Triah Joyce G. De la Cruz, University of the Philippines
Electrical and Electronics Engineering
Ramon Maria G. Acoymo, University of the Philippines
Voice and Music Theater/Dance Department,
Rowena Cristina L. Guevara, University of the Philippines
Electrical and Electronics Engineering Institute
Published
2010-09-28
Issue
Vol 21 No 2 (2009)
Section
Articles

Keywords

Filipino, Vowel Migration, Intelligibility
Submission of a manuscript implies: that the work described has not been published before (except in the form of an abstract or as part of a published lecture, review, or thesis); that it is not under consideration for publication elsewhere; that its publication has been approved by all co-authors, if any, as well as by the responsible authorities at the institute where the work has been carried out; that, if and when the manuscript is accepted for publication, the authors agree to the automatic transfer of the copyright to the publisher; that the manuscript will not be published elsewhere in any language without the consent of the copyright holders; that written permission of the copyright holder is obtained by the authors for material used from other copyrighted sources; and that any costs associated with obtaining this permission are the authors’ responsibility.

Most read articles by the same author(s)