Clarity logo
The 6th Clarity Workshop on Improving Speech-in-Noise for Hearing Devices (Clarity-2025)
ISCA logo

Results

Results of the 3rd Clarity Prediction Challenge are shown below for all submitted systems. For RMSE scores, small values are best. For correlation coefficients (corr), large numbers are best. The table also identifies which systems are intrusive and which are non-intrusive. The system labeled 'prior' simply returns the training set average sentence intelligibility for all sentences. The beHASPI system is the baseline supplied by the organisers.
Rank System Paper Intrusive? Dev RMSE Eval RMSE Eval Corr
E025 [paper] Yes 22.36 24.98 0.80
E019 [paper] No 21.87 25.31 0.79
E011a [paper] No 22.80 25.54 0.79
E020a [paper] Yes 23.15 25.60 0.78
E014 [paper] No 22.95 25.82 0.78
E011c [paper] No 22.89 25.83 0.79
E032 [paper] Yes (Text) 23.60 25.99 0.78
E020b [paper] Yes 23.47 26.02 0.78
E011b [paper] No 22.89 26.12 0.78
E020c [paper] Yes 24.81 26.14 0.77
E017 [paper] Yes 24.05 26.30 0.77
E024b [paper] No 24.74 26.58 0.77
E026 [paper] Yes 24.64 26.90 0.76
E039 [paper] Yes 25.61 27.00 0.76
E024a [paper] No 24.18 27.11 0.76
E038 [paper] Yes 24.88 27.82 0.74
E035 [paper] Yes 27.87 0.75
E012 [paper] No 26.28 29.07 0.75
HASPI Yes 28.00 29.47 0.70
E020c-ni [paper] No 30.16 32.74 0.65
E022a [paper] No 31.11 33.97 0.57
E022b [paper] No 33.10 35.48 0.56
Prior No 40.20 41.33

Prizes


The Hearing Industry Research Consortium best system prizes were awarded to the following teams.

First place prize, 1000 GPB
Hanlin Yu1, Haoshuai Zhou2, Linkai Li2,3, Boxuan Cao2, Changgeng Mo2, Shan Xiang Wang3 ( 1University of British Columbia,Canada; 2Orka Labs Inc., China; 3Stanford University, US), for their work:
Intrusive Intelligibility Prediction with ASR Encoders [Report]
Second place prize, 500 GPB
Rantu Buragohain1, Jejariya Ajaybhai1, Aashish Kumar Singh1, Karan Nathwani1, Sunil Kumar Kopparapu2 ( 1Indian Institute of Technology Jammu, India; 2Tata Consultancy Services Limited Mumbai, India), for their work:
Non-Intrusive Speech Intelligibility Prediction Using Whisper ASR and Wavelet Scattering Embeddings for Hearing-Impaired Individuals [Report]
Third place prize, 250 GPB
Ryandhimas E. Zezario1, Szu-Wei Fu2, Dyah A.M.G. Wisnu1,3, Hsin-Min Wang1, Yu Tsao1 ( 1Academia Sinica; 2NVIDIA; 3National Chengchi University), for their work:
Non-Intrusive Multi-Branch Speech Intelligibility Prediction using Multi-Stage Training [Report]

Congratulations to all our winners!


[TOP]