research
∙
04/04/2021
TSNAT: Two-Step Non-Autoregressvie Transformer Models for Speech Recognition
The autoregressive (AR) models, such as attention-based encoder-decoder ...
share
research
∙
03/17/2020
