Our mission at Karya does not stop at ethical data services
Our mission at Karya does not stop at ethical data services
Diverse Urdu read-speech corpus. Double validated. Includes lexicon
| Property | Value |
| Data Type | Monologue |
| Language | Urdu |
| Dialects | Dakhni, Dhakaiya, Rekhta, and Modern Vernacular Urdu |
| Lexicon Included | Yes |
| Sample Rate | 16 Hz |
| Contributors | 135 speakers, 45 mins per speaker |
| Accuracy | Minimum 95% |
This dataset is covered by our standard Karya Public License. Every purchase of this dataset provides royalties to it's contributors.