Back to open data
NLP & Language

DVoice — Moroccan Darija ASR Dataset

About

DVoice is an open source dataset for Automatic Speech Recognition (ASR) in Moroccan Darija. Contains voice recordings with text transcriptions. 2392 training files and 600 test files. Published by AIOXLABS, Zenodo 2021.

https://github.com/AIOXLABS/DVoice
Visit Website