Modeling speech recognition and synthesis simultaneously: Encoding and decoding lexical and sublexical semantic information into speech with no direct access to speech data.

Gasper Begus, Alan Zhou

Browse the full Interspeech paper archive.