Fused Acoustic and Text Encoding for
         Multimodal Bilingual Pretraining and Speech Translation

              Renjie Zheng * 1 Junkun Chen * 2 Mingbo Ma 1 Liang Huang 1 2

             Abstract
   Recently, representation learning for text and                           Abundant
                                              Speech
   speech has successfully improved many language
   related tasks. However, all existing methods suf-
   fer from two limitations: (a) they only learn from      ...