Skip to main content

The Clarity Project

About Us

About Us

The Clarity Project is a 5-year UKRI-funded research project involving four UK Universities and associated industrial partners. Our aim is to organise open evaluations for hearing aid algorithms and to champion novel machine learning approaches to hearing aid speech-in-noise processing.

Currently more than 1.5 billion people live with hearing loss, and this number is set to increase as the global population ages. Yet only a small fraction of people who could benefit from hearing aids have them, and most people who have the devices don't use them often enough. A major reason for this low uptake is the perception that hearing aids do not perform well enough.

We aim to stimulate progress in hearing aid signal processing by organising open machine learning challenges. These challenges will focus on speech-in-noise listening, a situation in which hearing aid users report the most dissatisfaction. For each challenge, we will be providing simulation tools, datasets and baseline systems. Our data and code will all be open-sourced, with the aim of lowering barriers that currently prevent speech and audio researchers from considering hearing impairment. Our funding is also allowing us to evaluate challenge submissions by running listening tests with hearing impaired listeners.

The Clarity Team

Michael Akeroyd
Michael AkeroydUniversity of Nottingham
Learn More
Jianyuan Sun
Jianyuan SunUniversity of Sheffield
Jon Barker
Jon BarkerUniversity of Sheffield
Trevor Cox
Trevor CoxUniversity of Salford
John Culling
John CullingUniversity of Cardiff
Simone Graetzer
Simone GraetzerUniversity of Salford
Graham Naylor
Graham NaylorUniversity of Nottingham
Zehai Tu
Zehai TuUniversity of Sheffield

The Clarity Steering Committee

Peter Derleth
Peter DerlethSonova AG
Christine Evers
Christine EversUniversity of Southampton
Ralph Holme
Ralph HolmeRoyal National Institute for the Deaf
Daniel Korzekwa
Daniel KorzekwaAmazon TTS Research

Our Clarity Partners

RNID logo.RNID logo.RNID logo.

Thanks to our funders

EPSRC logo.