Dataset Open Access

Dvoice : An open source dataset for Automatic Speech Recognition on African Languages and Dialects

Abdou Mohamed Naira; Anass Allak; Imade Benelallam

DVoice is a community initiative that aims to provide African languages and dialects with data and models to facilitate their use of voice technologies. The lack of data on these languages makes it necessary to collect data using methods that are specific to each language. Two different approaches are currently used: the DVoice platform, which is based on Mozilla Common Voice, for collecting authentic recordings from the community, and transfer learning techniques for automatically labeling the recordings. The DVoice platform currently manages 7 languages including Darija (Moroccan Arabic dialect) whose dataset appears on this version, Wolof, Mandingo, Serere, Pular, Diola and Soninke. The Swahili-labeled data present in this version was obtained after automatic labeling via the learning transfer of the Voxlingua107 dataset. For a first time, we also advocate for the increase of data given their small size that we currently have. Thus this version of the dataset contains easily identifiable augmented data.

Files (3.7 GB)
Name Size
v2.0.darija.tar.gz
md5:92592bff13cfa171d86b57127eb9035b
1.5 GB Download
v2.0.swahili.tar.gz
md5:e40e346499350d370aa49aa3491b7f46
2.2 GB Download
1,053
129
views
downloads
All versions This version
Views 1,0531,053
Downloads 129129
Data volume 232.4 GB232.4 GB
Unique views 1,0331,033
Unique downloads 5555

Share

Cite as