Clarity Challenges

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.