Consistent Transcription and Translation of Speech
| Autor: | Matthias Sperber, Christian Gollan, Matthias Paulik, Udhyakumar Nallasamy, Hendra Setiawan |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2020 |
| Předmět: |
FOS: Computer and information sciences
Linguistics and Language Computer science Inference 02 engineering and technology 010501 environmental sciences computer.software_genre 01 natural sciences User experience design Transcription (linguistics) Artificial Intelligence Speech translation 0202 electrical engineering electronic engineering information engineering 0105 earth and related environmental sciences Computer Science - Computation and Language business.industry Communication Strong consistency 020206 networking & telecommunications Computer Science Applications Human-Computer Interaction Artificial intelligence business computer Computation and Language (cs.CL) Natural language processing |
| Popis: | The conventional paradigm in speech translation starts with a speech recognition step to generate transcripts, followed by a translation step with the automatic transcripts as input. To address various shortcomings of this paradigm, recent work explores end-to-end trainable direct models that translate without transcribing. However, transcripts can be an indispensable output in practical applications, which often display transcripts alongside the translations to users. We make this common requirement explicit and explore the task of jointly transcribing and translating speech. While high accuracy of transcript and translation are crucial, even highly accurate systems can suffer from inconsistencies between both outputs that degrade the user experience. We introduce a methodology to evaluate consistency and compare several modeling approaches, including the traditional cascaded approach and end-to-end models. We find that direct models are poorly suited to the joint transcription/translation task, but that end-to-end models that feature a coupled inference procedure are able to achieve strong consistency. We further introduce simple techniques for directly optimizing for consistency, and analyze the resulting trade-offs between consistency, transcription accuracy, and translation accuracy. Accepted at TACL (pre-MIT Press publication version); added dataset link |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|