Abstract: Multilingual automatic speech recognition (ASR) models greatly facilitate recognizing low-resource languages by sharing representations across similar languages. However, the commonly ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback