Using public domain resources and off-the-shelf tools to produce high-quality multimedia texts
Date
2022
Authors
Rayner, M.
Chiera, B.
Chua, C.
Editors
Advisors
Journal Title
Journal ISSN
Volume Title
Type:
Conference paper
Citation
ALTA 2022, 2022, vol.20, pp.1-10
Statement of Responsibility
Conference Name
ALTA 2022 The 20th Annual Workshop of the Australasian Language Technology Association (14 Dec 2022 - 16 Dec 2022 : Adelaide, Australia)
Abstract
In the turbulent world of 2022, where mass population movements due to war and disaster are becoming increasingly common, language skills are more relevant than ever. People who wish to achieve a high level of proficiency when learning a new language benefit from reading literary texts, but many learners find this a challenging hurdle. Annotating texts with integrated audio and translations is a popular way to try and make them easier to approach. However, doing this automatically with TTS and machine translation engines produces unengaging results, while human annotation is slow and expensive. Here, we present a method that uses simple scripts and readily available computational resources for speech recognition and sentence alignment to combine public-domain resources from sites like Gutenberg and LibriVox into high-quality annotated multimedia versions of literary texts. Initial results with French texts of up to 80K words in length are promising, with audio/text word error rates under 0.25% and audio/translation word error rates around 1%, producing results that are usable after only minimal postediting.
School/Discipline
Dissertation Note
Provenance
Description
Access Status
Rights
Copyright 2022