Lecture by Researcher (Netherlands Cncer Institute), a visiting foreign faculty member (designated assistant professor) on June 17th, 2024．

Day and time: Monday, June 17, 2024, 10:30am-12:00pm
Venue: Lecture Room 113, Engineering Bldg. 1
Speaker: Bence Mark HALPERN (Researcher, Department of Head and neck Oncology, Netherlands Cancer Institute)
Topics: Pathological speech synthesis and evaluation
Language:  English Registration : Free Admission, no advance registration required 　　　

Abstract:
Pathological speech synthesis and evaluation is an emerging field within speech technology that focuses on the creation and analysis of speech signals affected by various disorders. This technology aims to develop tools and systems that can accurately replicate and analyse pathological speech patterns, providing significant benefits in fields such as speech therapy, assistive communication with automatic speech recognition. The lecture will cover terminologies that are required to understand pathological speech and disorders, and introduce the basic principles of the currently used state-of-the-art technologies for pathological speech synthesis and evaluation. We will discuss challenges, such as, evaluation the effectiveness and naturalness of synthesized pathological speech. This challenge remains a significant hurdle, as traditional evaluation metrics may not be suitable for assessing pathological speech. This necessitates the development of new evaluation frameworks that can account for the unique characteristics of disordered speech. References [1] Huang, W. C., Halpern, B. M., Violeta, L. P., Scharenborg, O., & Toda, T. (2022, May). Towards identity preserving normal to dysarthric voice conversion. In ICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) (pp. 6672-6676). IEEE. [2] Prananta, L., Halpern, B. M., Feng, S., & Scharenborg, O. (2022). The Effectiveness of Time Stretching for Enhancing Dysarthric Speech for Improved Dysarthric Speech Recognition. [3] Halpern, B. M., Feng, S., van Son, R., van den Brekel, M., & Scharenborg, O. (2022). Low-resource automatic speech recognition and error analyses of oral cancer speech. Speech Communication, 141, 14-27.

Biography:
Dr. Bence Halpern is actively engaged in research on information technology that supports healthcare. He has extensive experience in conducting field experiments in collaboration with medical sites and has developed technologies that are actually used in medical settings. In his doctoral dissertation, he focused on information technology to support speech therapy for individuals with speech impairments. His research has successfully produced various applied technologies, including speech recognition, automatic speech evaluation, and speech synthesis for people with speech impairments.