Ugiat/parlament_parla_v3_punctuated
Viewer • Updated • 332k • 556 • 3
Speech recognition with punctuation and capitals: datasets and models (Catalan, French).
Note 900 hours of Catalan parliamentary speech with punctuated and capitalized transcriptions.
Note 1,077 hours of French read speech (MLS) with punctuated and capitalized transcriptions.
Note Upstream: ParlamentParla v3 (BSC-LT, Aina project).
Note Upstream: Multilingual LibriSpeech.
Note Upstream: the MLS paper.