PhD defense by Paolo Mesiano

On Friday 19 August, Paolo Mesiano will defend his PhD thesis " Assessing the effects of fundamental-frequency dynamics on the intelligibility of competing voices".

 

Time: 14:00

Place: Building 101, Meeting Room M1 & zoom: https://dtudk.zoom.us/meeting/register/u5crf-usrjkiG9FdpZtU-M5XvET54fUeI8oq

Please be aware that the PhD defense may be recorded - This will also be informed at the beginning of the PhD defense.

 

Supervisor: Professor Torsten Dau
Co-Supervisor: Dr. Johannes Zaar

Dr. Helia Relaño-Iborra

 

Assessment committee:
Associate Professor Tobias May, DTU Health Tech

Professor Stuart Rosen, University College London (UCL)

Professor Emily Buss, University of North Carolina

 

Chairperson:
Senior Researcher Jens Bo Nielsen, DTU Health Tech

 

Abstract:
Hearing loss is a major obstacle in daily social life, as it can severely limit the participation in conversations, in particular when several interfering sound sources are present. Competing-talker scenarios are typical daily-life situation where a person is listening to the voice of a desired talker (i.e., the target) in the presence of other interfering voices. In such situations, people with a healthy auditory system can easily understand the target voice, while hearing-impaired people often have difficulties. In order to develop engineering solutions that can improve the hearing abilities of hearing-impaired people, it is necessary to understand what acoustic features of the speech sounds are useful for speech understanding in such challenging auditory situations.

This thesis focused on how the dynamics of the fundamental frequency of the voices in a competing-talker scenario can influence the intelligibility of the target speech. Three experiments were conducted on both normal-hearing and hearing-impaired listeners. It was shown that the fundamental frequency dynamics are useful for speech intelligibility in competing-talker scenarios, but hearing loss can severely limit their effects. Aspects of the fundamental frequency that may potentially improve speech intelligibility in hearing-impaired people are described.
Additionally, the fundamental frequency of speech recordings from Danish dialogues was analyzed, showing how the talkers, when in presence of communication barriers such as background noise or hearing loss, can adapt the fundamental frequency of their voice to promote the intelligibility of their speech for the interlocutor.
These findings offer an indication of how potential signal-processing manipulations of the speech signals targeted to the fundamental frequency can be implemented to improve speech intelligibility in hearing-impaired listeners.

Tidspunkt

fre 19 aug 22
14:00 - 17:00

Arrangør

Hvor

Building 101, Meeting Room M1 & zoom