STEMM: Self-learning with Speech-text Manifold Mixup for Speech Translation.

Qingkai Fang, Rong Ye, Lei Li, Yang Feng, Mingxuan Wang

Browse the full ACL paper archive.