Publication detail

Normalising Flows for Speaker and Language Recognition Backend

ESPUNA, A. PRASAD, A. MOTLÍČEK, P. MADIKERI, S. SCHUEPBACH, C.

Original Title

Normalising Flows for Speaker and Language Recognition Backend

Type

article in a collection out of WoS and Scopus

Language

English

Original Abstract

In this paper, we address the Gaussian distribution assumption made in PLDA, a popular back-end classifier used in Speaker and Language recognition tasks. We study normalizing flows, which allow using non-linear transformations and still obtain a model that can explicitly represent a probability density. The model makes no assumption about the distribution of the ob- servations. This alleviates the need for length normalization, a well known data preprocessing step used to boost PLDA performance. We demonstrate the effectiveness of this flow model on NIST SRE16, LRE17 and LRE22 datasets. We ob- serve that when applying length normalization, both the flow model and PLDA achieve similar EERs for SRE16 (11.5% vs 11.8%). However, when length normalization is not applied, the flow shows more robustness and offers better EERs (13.1% vs 17.1%). For LRE17 and LRE22, the best classification accu- racies (84.2%, 75.5%) are obtained by the flow model without any need for length normalization.

Keywords

Speaker recognition, Language Recognition

Authors

ESPUNA, A.; PRASAD, A.; MOTLÍČEK, P.; MADIKERI, S.; SCHUEPBACH, C.

Released

18. 6. 2024

Publisher

International Speech Communication Association

Location

Quebec

Pages from

74

Pages to

80

Pages count

7

URL

https://www.isca-archive.org/odyssey_2024/espuna24_odyssey.pdf

BibTex

@inproceedings{BUT193369,
  author="ESPUNA, A. and PRASAD, A. and MOTLÍČEK, P. and MADIKERI, S. and SCHUEPBACH, C.",
  title="Normalising Flows for Speaker and Language Recognition Backend",
  booktitle="Proceedings of Odyssey 2024: The Speaker and Language Recognition Workshop",
  year="2024",
  pages="74--80",
  publisher="International Speech Communication Association",
  address="Quebec",
  doi="10.21437/odyssey.2024-11",
  url="https://www.isca-archive.org/odyssey_2024/espuna24_odyssey.pdf"
}

Documents

espuna2024_odyssey.pdf