Clarity: Machine Learning Challenges to Revolutionise Hearing Device Processing

Simone Graetzer , Michael Akeroyd , Jon Barker (1) , Trevor Cox , John Culling (2) , Graham Naylor (3) , Eszter Porter

1 Department of Computer Science [Sheffield]
2 School of Psychology [Cardiff University]
3 Grasslands Research Centre

In the Clarity project, we will run a series of machine learning challenges to revolutionise speech processing for hearing devices. Over five years, there will be three paired challenges. Each pair will consist of a competition focussed on hearing-device processing and another focussed on speech perception modelling. The processing challenges will deliver new and improved approaches for hearing device signal processing for speech. The parallel perception challenges will develop and improve methods for predicting speech intelligibility and quality for hearing impaired listeners. To facilitate the challenges, we will generate open-access datasets, models and infrastructure. These will include: (1) tools for generating realistic test/training materials for different listening scenarios; (2) baseline models of hearing impairment; (3) baseline models of hearing-device processing; (4) baseline models of speech perception and (5) databases of speech perception in noise. The databases will include the results of listening tests that characterise how hearing-impaired listeners perceive speech in noise. We will also provide a comprehensive characterisation of each listener's hearing ability. The provision of open-access datasets, models and infrastructure will allow other researchers to develop their own algorithms for speech and hearing aid processing. In addition, it will lower barriers that prevent researchers from considering hearing impairment. In round one, speech will occur in the context of a living room, i.e., a moderately reverberant room with minimal (non-speech) background noise. Entries can be submitted to either the processing or perception challenges, or both. We expect to open round one in October 2020 for a closing date in June 2021 and results in October 2021. This EPSRC-funded project involves researchers from the Universities of Sheffield, Salford, Nottingham and Cardiff in conjunction with the Hearing Industry Research Consortium, Action on Hearing Loss, Amazon, and Honda. To register interest in the challenges, go to www.claritychallenge.org.

Mots clés: hearing; speech; machine learning; competition

Domaines: Acoustique [physics.class-ph]; Vibrations [physics.class-ph]

Simone Graetzer, Michael Akeroyd, Jon Barker, Trevor Cox, John Culling, et al.. Clarity: Machine Learning Challenges to Revolutionise Hearing Device Processing. Forum Acusticum, Dec 2020, Lyon, France. pp.3495-3497, ⟨10.48465/fa.2020.0198⟩. ⟨hal-03234191⟩