PI
Assistant Professor Jer-Ming Chen
Project Info
Speech and Singing are fundamental modes of human communication, expression, culture and social activity. Although voice science has developed extensively over the past several decades, the behavior of vocal tract geometry and its fine control during speech and singing still cannot be studied easily with good precision or in real time.
Consequently, studies of speech have so far been [largely] limited to the resulting speech sound produced (output), rather than the physical undergirding of how speech is produced and controlled (i.e. the vocal tract is often treated as a black box).
However, with the innovative application of novel acoustic techniques, combined with robust computation and machine learning models, we expect to shed light on vocal tract control parameters, voice acoustics, voice disorders, diagnostics, visualization, rehabilitation and training, thereby contributing directly to knowledge and an improved quality of life and social well-being for all.
Potential Research Scopes in Voice Science include
- Acoustic detection/classification of [chronic/acute] voice disorders and diagnosis
- Vocal tract parameter tracking in speech (phonology, prosody, accent)
- Broadband excitation of vocal tract towards speech diagnostics
- Rapid, non-invasive, estimations of vocal tract geometry via numerical and experimental methods (including developing actuated scale models)
- Acoustic and phonological survey of Asian languages and/or traditional singing modes, extending known acoustic models of voice production.
Interested students should contact Dr. Jer-Ming Chen (jerming_chen@sutd.edu.sg) with their recent CVs. Students are also encouraged to propose research projects according to their interests and abilities.