The CherISH network

Project 11_UK: Using artificial intelligence to optimise and maximise auditory training, specifically for bilateral cochlear implant users

Host Institution: Imperial college London (ICL), UK
Audio Experience Design (AXD) Research Group – Faculty of Engineering, Dyson School of Design Engineering

Background

It has been shown that sensory skills, specifically spatial hearing ones (e.g. the ability to precisely locate a sound source in the surrounding space), can be improved with training. This can be applied to someone learning to use significantly altered sound localisation cues, such as the ones provided by Cochlear Implants (CI). Virtual Reality (VR) has been extensively employed to implement specific training routines, which are though often designed using empirical approaches rather than in-depth research into the types of available tasks, the difficulty of the training, the individual circumstances of the participants, or again the individual performance during the training. Artificial Intelligence (AI) could be a very powerful tool for personalising and optimising training, as well as maximising outcomes.

Objectives
Design, implement and explore different VR-based spatial hearing training tasks and routines
Design and integrate an AI-based method for adaptively personalising and optimising the training for each specific participant
Use these tools to assess and quantify the effectiveness of the training, focusing specifically on bilateral CI users, and comparing the outcomes with training carried out using traditional non-adaptive methods.

Our Partners in this project:

• CBNL, Cochlear Benelux NV, Belgium [Secondment 1]
• UNITN, university of Trento, Italy [Secondment 2]