Publication Details
Speech Technology for Unwritten Languages
BESACIER, L.
BLACK, A.
Hasegawa-Johnson Mark (FIT)
Metze Florian
NEUBIG, G.
STÜKER, S.
GODARD, P.
MÜLLER, M.
ONDEL YANG, L.
PALASKAR, S.
ARTHUR, P.
CIANNELLA, F.
DU, M.
LARSEN, E.
MERKX, D.
RIAD, R.
WANG, L.
Dupoux Emmanuel (FIT)
Speech processing, automatic speech recognition, unsupervised learning, speech synthesis, image retrieval.
Abstract-Speech technology plays an important role in our everyday life. Among others, speech is used for human-computer interaction, for instance for information retrieval and on-line shopping. In the case of an unwritten language, however, speech technology is unfortunately difficult to create, because it cannot be created by the standard combination of pre-trained speech-to-text and text-to-speech subsystems. The research presented in this article takes the first steps towards speech technology for unwritten languages. Specifically, the aim of this work was 1) to learn speech-to-meaning representations without using text as an intermediate representation, and 2) to test the sufficiency of the learned representations to regenerate speech or translated text, or to retrieve images that depict the meaning of an utterance in an unwritten language. The results suggest that building systems that go directly from speech-to-meaning and from meaning-to-speech, bypassing the need for text, is possible.
@article{BUT170325,
author="SCHARENBORG, O. and BESACIER, L. and BLACK, A. and HASEGAWA-JOHNSON, M. and METZE, F. and NEUBIG, G. and STÜKER, S. and GODARD, P. and MÜLLER, M. and ONDEL YANG, L. and PALASKAR, S. and ARTHUR, P. and CIANNELLA, F. and DU, M. and LARSEN, E. and MERKX, D. and RIAD, R. and WANG, L. and DUPOUX, E.",
title="Speech Technology for Unwritten Languages",
journal="IEEE/ACM TRANSACTIONS ON AUDIO, SPEECH AND LANGUAGE PROCESSING",
year="2020",
volume="2020",
number="28",
pages="964--975",
doi="10.1109/TASLP.2020.2973896",
issn="2329-9290",
url="https://ieeexplore.ieee.org/document/8998182"
}