Datasets:
Tasks:
Automatic Speech Recognition
Modalities:
Audio
Formats:
soundfolder
Languages:
Spanish
Size:
10K - 100K
License:
Escucho Mucho — audio
Short Spanish speech clips (MP3, 24 kHz mono, ~64 kbps) used by the Escucho Mucho listening-practice app. Nothing here is original: the recordings are re-encoded copies of public speech corpora, republished so the app can stream them to a phone.
Each accent lives in its own folder; a clip's transcript, timings and difficulty live in the app's own library index, not in this repo.
| Folder | Source | Licence |
|---|---|---|
co/ |
OpenSLR SLR72 — Colombian Spanish | CC BY-SA 4.0 |
ar/ |
OpenSLR SLR61 — Argentinian Spanish | CC BY-SA 4.0 |
cl/ |
OpenSLR SLR71 — Chilean Spanish | CC BY-SA 4.0 |
pe/ |
OpenSLR SLR73 — Peruvian Spanish | CC BY-SA 4.0 |
pr/ |
OpenSLR SLR74 — Puerto Rican Spanish | CC BY-SA 4.0 |
ve/ |
OpenSLR SLR75 — Venezuelan Spanish | CC BY-SA 4.0 |
mx/ |
OpenSLR SLR39 — Heroico / USMA (Mexican Spanish) | CC BY 4.0 |
tedx/ |
TEDx Spanish corpus (Carlos Mena) | CC BY 4.0 |
Please credit the original corpora — see openslr.org for the crowdsourced Latin American Spanish datasets (Google / collaborators) and the TEDx Spanish corpus for the talk recordings. Share-alike terms of CC BY-SA 4.0 apply to the folders marked above.
- Downloads last month
- 253