TY  - RPRT
TI  - Multiple-hypothesis RNN-T Loss for Unsupervised Fine-tuning and Self-training of Neural Transducer
AU  - Cong-Thanh Do
AU  - Mohan Li
AU  - Rama Doddipatla
PY  - 2022
UR  - https://arxiv.org/abs/2207.14736
ID  - 2207.14736
ER  -