1 paper
Andrea Do, Oscar Brown, Zhengjie Wang +4
The performance of Whisper in low-resource languages is still far from perfect. In addition to a lack of training data on low-resource languages, we identify some limitations in th…