AdaST: Dynamically Adapting Encoder States in the Decoder for End-to-End Speech-to-Text Translation

Open in new window