Paul-Ambroise Duquenne

speech-to-speech translation (5) self-supervised learning (2) multilingual speech (2) multilingual translation (1) multimodal learning (1) parallel corpus (1) automatic speech recognition (1) cross-modal transfer (1) mixture of expert (1) multimodal representation (1) end-to-end training (1) evaluation metric (1) model pre-training (1) discrete representation (1) end-to-end translation (1) human evaluation (1) cosine similarity (1) modular architecture (1) multilingual training (1) multimodal embedding (1)