(Translated by https://www.hiragana.jp/)
[2210.11642] Improving Semi-supervised End-to-end Automatic Speech Recognition using CycleGAN and Inter-domain Losses