Michael Kai Petersen

Senior Researcher

> Learn more


The needs of hearing aid users change according to the environment and listening situation they find themselves in. We track data to identify these needs in order to crack the code to personalizing hearing aids and, eventually, make better hearing solutions for the individual.

A hearing aid fitted just for you

All hearing aid users are different, and so are the requirements to their hearing aids, dependent on the listening situations. Sometimes the user may need clear crisp voices, in other situations they may wish to hear every single sound around them, and sometimes the background noise needs to be reduced. The aim of this focus area is to identify ways to personalize the listening experience in order to create hearing aids that will be able to choose the best hearing aid settings itself, based on the user’s learned preferences in multiple contexts.

How can we understand the need of the user based on data?

We are using data on the hearing aid users’ choice of programme in specific contexts to learn the preferences of the user. The hearing aids gather the data, to infer which programme the test subject prefers in specific situations by including additional context parameters related to motion, location or activity. In order for hearing aids to learn behavioural patterns and adapt to the context, the user needs to be an active part of the solution.

Associated PhD students

Benjamin Johansen, who is a PhD student at Cognitive Systems Section at DTU Compute, Technical University of Denmark, and The CopenhAgen Center for HEalth Technology (Cachet). Read more about him here.

Maciej Korzepa, who is a PhD student at DTU Compute, Technical University of Denmark.

Further reading

Johansen B, Korzepa M, Petersen MK, Pontoppidan NH, Eg Larsen J (2018). Mapping auditory percepts into visual interfaces for hearing impaired users. Proceedings of Human Factors in Computing Systems (CHI 2018)

Korzepa M, Johansen B, Petersen MK, Pontoppidan NH, Larsen J, Eg Larsen J (2018). Learning preferences and soundscapes for augmented hearing.  Proceedings of Intelligent User Interfaces (IUI 2018)

Johansen B, Korzepa M, Petersen MK, Pontoppidan NH, Larsen J, Eg Larsen J (2017). Personalizing the fitting of hearing aids by learning contextual preferences from internet of things data.  MDPI Computers 7:1, 2017 http://dx.doi.org/10.3390/computers7010001

Johansen B, Petersen MK, Pontoppidan NH, Sandholm P, Eg Larsen J (2017). Rethinking hearing aid fitting by learning from behavioral patterns. Proceedings of Human Factors in Computing Systems (CHI 2017)