SLaTE 2015 - Workshop on Speech and Language Technology in Education

Leipzig, Germany
September 4-5, 2015

Development of a Prosodic Reading Tutor of Japanese – Effective Use of TTS and F0 Contour Modeling Techniques for CALL

Nobuaki Minematsu (1), Hiroya Hashimoto (1), Hiroko Hirano (2), Daisuke Saito (1)

(1) The University of Tokyo, Tokyo, Japan
(2) Tokyo University of Foreign Studies, Tokyo, Japan

A text typed to a speech synthesizer is generally converted into its corresponding phoneme sequence on which various kinds of prosodic symbols are attached by a prosody prediction module. By using this module effectively, we build a prosodic reading tutor of Japanese, called Suzuki-kun, and it is provided as one function of OJAD (Online Japanese Accent Dictionary). In Suzuki-kun, by using a prosody prediction module, any Japanese text is converted into its reading (Hiragana sequence) on which the pitch pattern that sounds natural is visualized as a smooth curve drawn by the F0 contour generation process model. Further, positions of accent nuclei and unvoiced vowels are illustrated. Suzuki-kun also reads that text out following the prosodic features that are visualized. Since releasing Suzuki-kun, the number of accesses to OJAD has been drastically increased and for the last four months, OJAD received 129,168 accesses, 58.9% of which were from outside Japan.

Full Paper

Bibliographic reference.  Minematsu, Nobuaki / Hashimoto, Hiroya / Hirano, Hiroko / Saito, Daisuke (2015): "Development of a prosodic reading tutor of Japanese – effective use of TTS and F0 contour modeling techniques for CALL", In SLaTE-2015, 189.