Direct speech translation (ST) has shown to be a complex task requiring knowledge transfer from its sub-tasks: automatic speech recognition (ASR) and machine translation (MT). For MT, one of the most promising techniques to transfer knowledge is knowledge distillation. In this paper, we compare the different solutions to distill knowledge in a sequence-to-sequence task like ST. Moreover, we analyze eventual drawbacks of this approach and how to alleviate them maintaining the benefits in terms of translation quality.
On knowledge distillation for direct speech translation / Gaido, M.; Di Gangi, M. A.; Negri, M.; Turchi, M.. - 2769:(2020). (Intervento presentato al convegno 7th Italian Conference on Computational Linguistics, CLiC-it 2020 tenutosi a Bologna, Italy nel 1-3 March, 2021).
On knowledge distillation for direct speech translation
Gaido M.;Di Gangi M. A.;
2020-01-01
Abstract
Direct speech translation (ST) has shown to be a complex task requiring knowledge transfer from its sub-tasks: automatic speech recognition (ASR) and machine translation (MT). For MT, one of the most promising techniques to transfer knowledge is knowledge distillation. In this paper, we compare the different solutions to distill knowledge in a sequence-to-sequence task like ST. Moreover, we analyze eventual drawbacks of this approach and how to alleviate them maintaining the benefits in terms of translation quality.| File | Dimensione | Formato | |
|---|---|---|---|
|
paper_28.pdf
accesso aperto
Descrizione: Articolo principale
Tipologia:
Versione editoriale (Publisher’s layout)
Licenza:
Creative commons
Dimensione
301.27 kB
Formato
Adobe PDF
|
301.27 kB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione
