Published September 27, 2024
| Version v1
Model
Open
SCOREQ: Speech Quality Assessment with Contrastive Regression - PyTorch weights
Creators
-
1.
University College Dublin
- 2. Insight Centre for Data Analytics
Description
| Model | Training Set/Domain | Usage Mode | Margin |
| adapt_nr_telephone.pt | NISQA TRAIN SIM / Telephone speech | No-reference | Adaptive |
| fixed_nmr_telephone.pt | NISQA TRAIN SIM / Telephone speech | Non-matching reference, full-reference | Constant |
| adapt_nr_synthetic.pt | VoiceMOS 22 / Synthetic speech | No-reference | Adaptive |
| fixed_nmr_synthetic.pt | VoiceMOS 22 / Synthetic speech | Non-matching reference | Constant |
Notes
Cite as
Alessandro Ragano, Jan Skoglund, and Andrew Hines (2024) SCOREQ: Speech Quality Assessment with Contrastive Regression, Advances in Neural Information Processing Systems (NeurIPS) 2024.
Files
Files
(1.5 GB)
| Name | Size | Download all |
|---|---|---|
|
md5:3c9c121f1b9e568de8616911acc98e02
|
378.4 MB | Download |
|
md5:87811369e1bbba353ee6cc560d5c17ac
|
378.4 MB | Download |
|
md5:5f12a7eefdbbe1db52363b3495934441
|
378.4 MB | Download |
|
md5:6157e47bdbeda0e0e3b7a2b0c868f272
|
378.4 MB | Download |