Publication Details
A Hierarchical Subspace Model for Language-Attuned Acoustic Unit Discovery
ONDEL YANG, L.
Burget Lukáš, doc. Ing., Ph.D. (DCGM)
Černocký Jan, prof. Dr. Ing. (DCGM)
SARAÇLAR, M.
acoustic unit discovery, hierarchical subspace model, unsupervised learning
In this work, we propose a hierarchical subspace model for acoustic unit
discovery. In this approach, we frame the task as one of learning embeddings on
a low-dimensional phonetic subspace, and simultaneously specify the subspace
itself as an embedding on a hyper- subspace. We train the hyper-subspace on a set
of transcribed languages and transfer it to the target language. In the target
language, we infer both the language and unit embeddings in an unsupervised
manner, and in so doing, we simultaneously learn a subspace of units specific to
that language and the units that dwell on it. We conduct experiments on TIMIT and
two low-resource languages: Mboshi and Yoruba. Results show that our model
outperforms major acoustic unit discovery techniques, both in terms of clustering
quality and segmentation accuracy.
@inproceedings{BUT175792,
author="YUSUF, B. and ONDEL YANG, L. and BURGET, L. and ČERNOCKÝ, J. and SARAÇLAR, M.",
title="A Hierarchical Subspace Model for Language-Attuned Acoustic Unit Discovery",
booktitle="ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)",
year="2021",
pages="3710--3714",
publisher="IEEE Signal Processing Society",
address="Toronto, Ontario",
doi="10.1109/ICASSP39728.2021.9414899",
isbn="978-1-7281-7605-5",
url="https://www.fit.vut.cz/research/publication/12523/"
}