Speech recognition system trains on radio archive to learn Niger Congo languages

Researchers at Stanford and GNCode created a new dataset and model trained from radio archives of ‘low-resource’ African languages.Read More