DagsHub / audio-datasets

open-source audio datasets
https://dagshub.com/DagsHub/audio-datasets
141 stars 23 forks source link

Free spoken digit dataset #49

Closed kinkusuma closed 3 years ago

kinkusuma commented 3 years ago

Dataset: FSDD
About: 4 speakers, 2,000 recordings (50 of each digit per speaker), English pronunciations.