Computer-assisted learning of English pronunciation and speaking

Activity: Talk or presentationInvited talk

Description

The articulation of Japanese and English sounds involves different muscle memories and coordination, so frequent articulatory practice is essential. This two-part workshop is designed for Japanese learners of English as well as engineers interested in developing computer-assisted and/or mobile pronunciation training apps. In part 1, I will provide an overview of useful tools for facilitating English pronunciation practice and discuss the technologies underlying the tools (e.g. automatic speech recognition, dialogue systems). I will cite relevant research to discuss the pros and cons of pronunciation training with a teacher versus a computer. I will also highlight how recent advances in speech technologies such as prosodic transplantation and voice-cloning may facilitate improvements in L2 learners’ understanding and use of tone of voice. Part 2 will focus on speech prosody, illustrating how the nine tones in English convey different attitudinal meanings and the discourse, psycholinguistic and structural functions of speech prosody. There will be hands-on English pronunciation practice using various apps, including English Central, Google Translate, Siri and Resemble.ai. Finally, I will present my new machine-learning project which aims at providing automatic feedback on the accuracy of users’ imitation of the tone of voice of presenters heard on YouTube videos.
Period1 Nov 2022
Held atUniversity of Tsukuba, Japan
Degree of RecognitionInternational