2nd Clarity Prediction Challenge (CPC2) dataset for hearing aid speech intelligibility prediction

This dataset was created for the second Clarity Prediction Challenge (CPC2), which investigated the prediction of speech intelligibility for hearing-aid users in noisy acoustic environments. The challenge is now complete, but the dataset is made available for ongoing research. It can be used to develop and evaluate algorithms that estimate speech intelligibility across a range of acoustic and hearing-aid conditions.

The release includes audio signals, listener metadata, and supporting documentation. These materials enable reproducible evaluation of past challenge submissions as well as benchmarking of new approaches.

Further information about the challenge, including background, task definitions, and baseline systems, is available at: https://claritychallenge.org/docs/cpc2/cpc2_intro

How to cite: If this dataset is used in published work, please cite both this Zenodo record and the associated conference paper describing the dataset. The paper provides the main technical description of the data, including its collection protocol, structure, and intended research use.

Barker, J., Akeroyd, M. A., Bailey, W., Cox, T. J., Culling, J. F., Firth, J., Graetzer, S., & Naylor, G. (2024). The 2nd Clarity Prediction Challenge: A Machine Learning Challenge for Hearing Aid Intelligibility Prediction. In ICASSP 2024 - 2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Seoul, Korea, Republic of, pp. 11551–11555. IEEE. https://doi.org/10.1109/ICASSP48485.2024.10446441