- Position identifier: ESR14
- Host partner: Fraunhofer IDMT
Objectives
In this project, we aimed to evaluate the efficacy of electroencephalography (EEG) as an objective measure of listening effort. Identifying reliable measures of listening effort which reflect the experience is very important, as prolonged effort induced by noisy environments or hearing impairment can lead to listener discomfort, fatigue, and a reduced quality of life (Hetu et al., 1988; Hicks & Tharpe, 2002). Many approaches exist to enhance speech for playback in noisy conditions and many significantly improve speech intelligibility, but less is known about how they can relieve listening effort. This is particularly important as listening effort has been found to be sensitive to signal changes even when speech intelligibility is at 100% correct. Several methods of assessing listening effort exist, including subjective self-reports, behavioural tasks, pupillometry, and neurophysiological recordings, but the relative sensitivity of these requires further investigation. Project 14 therefore compared EEG with other methods for assessing the relief of listening effort provided by speech enhancement methods.
I am currently on secondment in Edinburgh, running my second experiment. Here I am drawing on the knowledge of Avashna Govender and Carol Chermaz to compare the efficacy of EEG and pupillometry as measures of listening effort. The stimuli are varied in terms of cognitive load by implementing different signal-to-noise ratios and the speech enhancement algorithm, AdaptDRC.
This project was supervised by Dr. Axel Winneke and Jan Rennies-Hochmuth at Fraunhofer IDMT, with collaboration by Dr. Volker Hohmann from Carl von Ossietzky Universität Oldenburg.