ASR Posterior-Based Loss for Multi-Task End-to-End Speech Translation.

Yuka Ko, Katsuhito Sudoh, Sakriani Sakti, Satoshi Nakamura

Browse the full Interspeech paper archive.