Update README.md
Browse files
README.md
CHANGED
|
@@ -58,7 +58,6 @@ The quality scores are distilled from two established audio quality models:
|
|
| 58 |
| **Background Quality** | Quality of the background (absence of noise/artifacts). Higher is better. | DNSMOS | 1.0 - 4.3 | ~3.2 | MOS-like |
|
| 59 |
| **Content Enjoyment** | How engaging/enjoyable the spoken content is. Higher is better. | Meta AudioBox | 1.9 - 5.1 | ~4.1 | MOS-like |
|
| 60 |
|
| 61 |
-
These scores were trained on the [mitermix/balanced-audio-score-datasets](https://huggingface.co/datasets/mitermix/balanced-audio-score-datasets) dataset.
|
| 62 |
|
| 63 |
### Validation Performance
|
| 64 |
|
|
|
|
| 58 |
| **Background Quality** | Quality of the background (absence of noise/artifacts). Higher is better. | DNSMOS | 1.0 - 4.3 | ~3.2 | MOS-like |
|
| 59 |
| **Content Enjoyment** | How engaging/enjoyable the spoken content is. Higher is better. | Meta AudioBox | 1.9 - 5.1 | ~4.1 | MOS-like |
|
| 60 |
|
|
|
|
| 61 |
|
| 62 |
### Validation Performance
|
| 63 |
|