Conference paper
Predicting Speech Intelligibility based on Fluctuations in Simulated Auditory-Nerve Responses
Abstract
Various speech intelligibility (SI) models have been proposed to predict the ability of normal-hearing (NH) listeners to understand speech in adverse listening conditions. However, most current SI models are based on a strongly simplified linear simulation of the highly non-linear auditory periphery, which limits their ability to predict effects of hearing impairment on SI.
At the same time, the models’ decision stages typically interact strongly with the type of auditory front-end processing applied.
| Language: | English |
|---|---|
| Year: | 2018 |
| Proceedings: | 41st Midwinter Meeting of the Association for Research in Otolaryngology |
| Types: | Conference paper |
| ORCIDs: | Dau, Torsten |
