Contact person :
fethi.bougares@elyadata.com
We introduce TUNIFRA, a novel and comprehensive corpus developed to advance research in Automatic Speech Recognition (ASR)
and Speech-to-Text Translation (STT) for Tunisian Arabic, a notably low-resourced language variety.
The TUNIFRA corpus comprises 15 hours of native Tunisian Arabic speech, carefully transcribed and manually… See the full description on the dataset page:
https://huggingface.co/datasets/fbougares/TUNIFRA.
