Skip to main content

3 posts tagged with "CPC2"

View All Tags

CPC2 eval data released

· One min read
Jon Barker
Clarity Team Member

The CPC2 evaluation data has now been released.

The data is available for download as a single 478 MB file, clarity_CPC2_data.test.v1_0.tgz. The evaluation data should be untarred into the same root as the training data. Further details can be found on the challenge website.

The data consists of the hearing aid algorithm output signals, clean reference signals, listener audiograms, and head rotation information. Listener responses are not provided for the evaluation data but will be made available after the submission window has closed.

For details on how to prepare your submission please see the instructions on the website.

If you have any questions please feel free to post them on this forum.

The submission window will close on the 31st of July.

Good luck!

Clarity-2023 Workshop @ Interspeech, Dublin

· 3 min read
Jon Barker
Clarity Team Member

We are pleased to announce the 4th ISCA Clarity Workshop on Machine Learning Challenges for Hearing Aids (Clarity-2023).

The event will be a one-day workshop held as an ISCA satellite event to Interspeech 2023 in Dublin, Ireland.

For registration and programme details please visit the workshop website


  • 2nd June 2023 - Workshop Submission Deadline (Regular Papers)
  • 31st July 2023 - Workshop Submission Deadline (Clarity Challenge Papers)
  • 5th August 2023 - Registration closes
  • 19th August - Workshop / Clarity Challenge results announced


One of the biggest challenges for hearing-impaired listeners is understanding speech in the presence of background noise. Everyday social noise levels can have a devastating impact on speech intelligibility. The inability to communicate effectively can lead to social withdrawal and isolation. Disabling hearing impairment affects 360 million people worldwide, with that number increasing because of the ageing population. Unfortunately, current hearing aid technology is often ineffective in noisy situations. Although amplification can restore audibility, it does not compensate fully for the effects of hearing loss.

The Clarity workshops are designed to stimulate a two-way conversation between the speech research community and hearing aid developers. Hearing aid developers, who are not typically represented at Interspeech, will have an opportunity to present the challenges of their industry to the speech community; the speech community will be able to present and discuss potentially transformative approaches to speech in noise processing in the presence of hearing researchers and industry experts.


Any work related to the challenges of hearing aid signal processing will be considered relevant topics include,

  • Binaural technology for speech enhancement and source separation
  • Multi-microphone processing technology
  • Real-time approaches to speech enhancement
  • Statistical model-driven approaches to hearing aid processing
  • Audio quality & intelligibility assessment hearing aid and cochlear implant users
  • Efficient and effective integration of psychoacoustic testing in machine learning
  • Machine learning for diverse target listeners
  • Machine learning models of hearing impairment

The 2nd Clarity Prediction Challenge

The Clarity-2023 will also host the 2nd Clarity Prediction Challenge, that is addressing the problem of developing new intrusive and non-intrusive approaches to hearing-aid speech intelligibility prediction. The Challenge will be launching on 1st March, is you may be interested in participating please sign up to our Google group for further announcements.

Keynote Talks

  • Prof Fei Chen, SUSTech, China,
  • Prof DeLiang Wang, Ohio State University, US


  • Michael Akeroyd, University of Nottingham
  • Jon Barker, University of Sheffield
  • Trevor Cox, University of Salford
  • Fei Chen, Southern University of Science and Technology, China
  • John Culling, University of Cardiff
  • Simone Graetzer, University of Salford
  • Andrew Hines, University College Dublin

For further information

To be kept up to date please join our Clarity Challenge Google group. If you have questions, please contact us directly using the contact details found here.

Funded by the Engineering and Physical Sciences Research Council (EPSRC), UK

Supported by RNID (formerly Action on Hearing Loss), Hearing Industry Research Consortium, Amazon TTS Research

Announcing the 2nd Clarity Prediction Challenge (CPC2)

· 2 min read
Jon Barker
Clarity Team Member
Trevor Cox
Clarity Team Member

The 2nd Clarity Prediction Challenge - Register Now

To allow the development of better hearing aids, we need ways to evaluate the speech intelligibility of audio signals automatically. We need a prediction model that takes the audio produced by a hearing aid and the listener's characteristics (e.g. audiogram) and estimates the speech intelligibility score that the listener would achieve in a listening test.

Last year we ran the CPC1 Challenge to develop such models. The challenge was presented at an online workshop and a special session of Interspeech 2022. We are now running the 2nd round of this challenge (CPC2), which builds on the first by using more complex signals and a larger set of listening test data for training and evaluating the prediction systems.

The outputs of the new challenge will be presented at an ISCA workshop that is being run as a satellite event to Interspeech 2023 in Dublin on 19th August 2023.

Full details can be found on the Challenge Website.

Register now to take part

If you are interested in participating please register now via the online registration form.

Important Dates

  • March - Launch of challenge, release of training data + baseline system.
  • 1st July - Release of evaluation data and opening of submission window.
  • 31st July - Submission deadline.
  • 19th August - ISCA Clarity 2023 workshop @ Interspeech
  • 19th September - Deadline for submission of finalised Workshop papers

What will be provided

  • Audio produced by a variety of (simulated) hearing aids for speech-in-noise;
  • The corresponding clean reference signals (the original speech);
  • Characteristics of the listeners (pure tone audiograms, etc);
  • The measured speech intelligibility scores from listening tests, where hearing-impaired listeners were asked to say what they heard after listening to the hearing aid processed signals.
  • Software tools including a baseline system based on HASPI scores.

For further information

To be kept up to date please join our Clarity Challenge Google group. If you have questions, please contact us directly using the contact details found here.