A hierarchical subspace model for language-attuned acoustic unit discovery
2021 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2021, Virtual, Toronto, Kanada, 6 - 11 Haziran 2021, cilt.2021-June, ss.3710-3714, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Cilt numarası: 2021-June
- Doi Numarası: 10.1109/icassp39728.2021.9414899
- Basıldığı Şehir: Virtual, Toronto
- Basıldığı Ülke: Kanada
- Sayfa Sayıları: ss.3710-3714
- Anahtar Kelimeler: acoustic unit discovery, hierarchical subspace model, unsupervised learning
- Açık Arşiv Koleksiyonu: AVESİS Açık Erişim Koleksiyonu
- Boğaziçi Üniversitesi Adresli: Evet
Özet
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.