Sign inGet started
Dataset · Hugging Face

IsmatillayevMuzaffar/uzbek-stt-synthetic

Uzbek STT — synthetic speech 103 hours, 52,460 clips of synthetic Uzbek speech for training and evaluating speech recognition. Grows over time (synthesized in 100 h stages). ⚠️ Synthetic data. Audio is generated by OmniVoice (Apache-2.0). ~80% of clips use zero-shot voice cloning from reference clips of openly licensed speech corpora (UzbekVoice — apache-2.0, USC/ISSAI — CC-BY-4.0, Common Voice — CC0); ~20% use text-described voices. Intended only for ASR research and training… See the full description on the dataset page: https://huggingface.co/datasets/IsmatillayevMuzaffar/uzbek-stt-syntheti

Author / organizationIsmatillayevMuzaffar
License (as declared)cc-by-sa-4.0
Downloads291 (as reported by Hugging Face)
Likes1
Created7 October 2026
Last updated at source18 October 2026
syntheticspeechuzbek
Embed this entry
<iframe src="https://silytics.si/embed/item/fbf4df64-1f0b-42e9-9993-644db978db37" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Popularity, last 3 days

291 now (+86 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.

Recent changes

  • 18 Oct 2026 Description: Uzbek STT — synthetic speech 77 hours, 39,702 clips of synthetic Uzbek speech for training and evaluating speech recognition. Grows over time (synthesized in 100 h stages). ⚠️ Synthetic data. Audio is generated by OmniVoice (Apache-2.0). ~8 → Uzbek STT — synthetic speech 103 hours, 52,460 clips of synthetic Uzbek speech for training and evaluating speech recognition. Grows over time (synthesized in 100 h stages). ⚠️ Synthetic data. Audio is generated by OmniVoice (Apache-2.0). ~
  • 17 Oct 2026 Description: Uzbek STT — synthetic speech 46 hours, 23,861 clips of synthetic Uzbek speech for training and evaluating speech recognition. Grows over time (synthesized in 100 h stages). ⚠️ Synthetic data. Audio is generated by OmniVoice (Apache-2.0). ~8 → Uzbek STT — synthetic speech 77 hours, 39,702 clips of synthetic Uzbek speech for training and evaluating speech recognition. Grows over time (synthesized in 100 h stages). ⚠️ Synthetic data. Audio is generated by OmniVoice (Apache-2.0). ~8

Changes detected by Silytics between refreshes of the source.

More from IsmatillayevMuzaffar

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.