Models for hearing devices
Since 1 April, 2018, the former head of the group holds a professorship at Lübeck University of Applied Sciences.
The group's project work is currently being continued in Oldenburg.
The fitting of hearing devices (cochlear implants and hearing aids) today to the user is done by taking only a few audiological measurement results into account, such as the individual absolute threshold of hearing. Details about the performance of the users in specific hearing tasks (such as speech in noise tests) are often tested after the fitting, but are not used for optimizing the algorithm fitting.
In contrast, the research paradigm of the group “Models for hearing devices” is to use auditory models that replicate the signal processing in the individual listener and tune hearing devices according to the responses of the model. Such “artificial listeners” are created both individually and as group averages. Such a paradigm has basically two advantages:
1. It is possible to learn about the interaction of the signal processing in the listener (mimicked by the model) and the signal processing of the hearing device. This helps to understand, e.g., what signal processing purpose certain physiologic aspects of the auditory system can have.
2. It offers the capability to test (and tune) different hearing device algorithms with the model without complications such as training effects and fatigue.
Afterwards, an optimal hearing device can be given to the respective hearing-impaired listener.
The principal goal is to optimize and improve the signal processing in the device to the specific listener, but we are also interested in improving our auditory models. Furthermore, we need to test if the predicted benefits with the algorithms can also be seen in the listeners. Therefore, we also test participants with normal hearing, hearing impairment or implanted listeners with psychoacoustic tests or speech-in-noise tests, such as the Oldenburg sentence test.
This lab is integrated in the Cluster of excellence “Hearing4all” (link) in task group 4 “Improving signal processing of hearing devices”. With our model-based approach, we are complementing the work of the groups “auditory scene analysis for hearing devices (aHeaD)” and “Individualized hearing devices (iHeaD)” with the common goal of improving hearing devices with all available techniques.
We are constantly looking for research assistants and students, who are interested in doing their BSc, MSc, or PhD-thesis with us. Possible topics include measurement-focused, signal-processing-focused and model-focused topics.
Further information about each of the topics: tim.juergens(at)uni-oldenburg.de