Speech and Language Technology in Education (SLaTE 2013)

Grenoble, France
August 30-September 1, 2013

Naturalness on Japanese Pronunciation before and after Shadowing Training and Prosody Modified Stimuli

A. Rongna (1), Ryoko Hayashi (2), Tatsuya Kitamura (3)

(1) Department of Rehabilitation for Sensory Functions, Research Institute of National Rehabilitation Center for Persons with Disabilities, Tokorozawa, Japan
(2) Graduate School of Intercultural Studies, Kobe University, Kobe, Japan
(3) Faculty of Intelligence and Informatics, Konan University, Kobe, Japan

This study attempts to investigate the change of naturalness impression for the Japanese utterance by Japanese as foreign language learners (JFL) before and after shadowing and repeating trainings and to discuss the crucial prosodic cue for the naturalness judgment.
   The speech of 8 JFL learners before and after pronunciation training was used, and their durational pattern, pitch pattern, or both of them were replaced with those of the model speech. 52 Japanese native speakers (JNS) assessed the naturalness of these stimuli. The results showed JNS judge the duration AND F0 modified stimuli most natural. In addition, the shadowing trained group tended to get higher scores than the repeating trained group after training. The acoustical analysis of speech material showed a difference of moraic structure and pitch accent between the shadowing and repeating group.

Index Terms: Japanese as a foreign language learners, naturalness, shadowing, repeating, prosodic modified stimuli

Full Paper

Bibliographic reference.  Rongna, A. / Hayashi, Ryoko / Kitamura, Tatsuya (2013): "Naturalness on Japanese pronunciation before and after shadowing training and prosody modified stimuli", In SLaTE-2013, 143-146.