by Simon Carlile

The biennial IHCON conference was held in smokey Banff from August 5th to 9th and attended by a number of AFHI Collaborators. The conference is a single track conference with time divided about equally between the podium and the poster presentations. AFHI had two podium presentations and two poster presentations spread over the 4 days.

The Machine Learning Hearing Aid work stream had a poster describing our consonant confusion test together with the ML based consonant classifier and a description of how this was being used in training a computational model of different kinds of cochlear dysfunction.

Dick Lyon had a podium presentation where he described how the cochlear model (CARFAC) was integrated into the machine learning training loop and shared some early results comparing how it can be used to fit a conventional hearing with fitting predictions from NAL-NL2 and DSL v5.0.

Both the poster and the podium presentations were very well attended and generated much discussion and interaction. Another highlight of the conference was the chance to have a close look at what other research groups were doing in this area as well as progress discussion on how we can collaborate more closely. Since the conference AFHI has already started collaborating with Professor Hari Bharadwaj and his group from the University of Pittsburg on different approaches to modelling the cochlear dysfunctions underlying hearing using the CARFAC model. Other collaborative links included Professor Sarah Verhulst from Ghent in Belgium as well as with research groups in Oldenburg (Germany) and UCL (London).

See the accompanying blog post regarding Kelly’s presentations about the PRISM EMA system and the importance of hearing testing in stroke.