Publication Details

Training Speaker Embedding Extractors Using Multi-Speaker Audio with Unknown Speaker Boundaries

STAFYLAKIS, T.; MOŠNER, L.; PLCHOT, O.; ROHDIN, J.; SILNOVA, A.; BURGET, L.; ČERNOCKÝ, J. Training Speaker Embedding Extractors Using Multi-Speaker Audio with Unknown Speaker Boundaries. In Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH. Proceedings of Interspeech. Incheon: International Speech Communication Association, 2022. p. 605-609. ISSN: 1990-9772.

Czech title

Trénování extraktorů embeddingů mluvčích pro nahrávky s několika mluvčími s neznámou segmentací

Type

conference paper

Language

English

Authors

Stafylakis Themos
Mošner Ladislav, Ing. (DCGM)
Plchot Oldřich, Ing., Ph.D. (DCGM)
Rohdin Johan Andréas, M.Sc., Ph.D. (DCGM)
Silnova Anna, M.Sc., Ph.D. (DCGM)
Burget Lukáš, doc. Ing., Ph.D. (DCGM)
Černocký Jan, prof. Dr. Ing. (DCGM)

URL

Keywords

Speaker Embedding Extractors, Multi-Speaker Audio, Unknown Speaker Boundaries

Abstract

In this paper, we demonstrate a method for training speaker em- bedding extractors using weak annotation. More specifically, we are using the full VoxCeleb recordings and the name of the celebrities appearing on each video without knowledge of the time intervals the celebrities appear in the video. We show that by combining a baseline speaker diarization algorithm that re- quires no training or parameter tuning, a modified loss with aggregation over segments, and a two-stage training approach, we are able to train a competitive ResNet-based embedding extractor. Finally, we experiment with two different aggregation functions and analyze their behaviour in terms of their gradients.

Published

2022

Pages

605–609

Journal

Proceedings of Interspeech, vol. 2022, no. 9, ISSN 1990-9772

Proceedings

Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH

Publisher

International Speech Communication Association

Place

Incheon

DOI

10.21437/Interspeech.2022-10165

UT WoS

000900724500123

EID Scopus

2-s2.0-85140088494

BibTeX

@inproceedings{BUT179781,
  author="Themos {Stafylakis} and Ladislav {Mošner} and Oldřich {Plchot} and Johan Andréas {Rohdin} and Anna {Silnova} and Lukáš {Burget} and Jan {Černocký}",
  title="Training Speaker Embedding Extractors Using Multi-Speaker Audio with Unknown Speaker Boundaries",
  booktitle="Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH",
  year="2022",
  journal="Proceedings of Interspeech",
  volume="2022",
  number="9",
  pages="605--609",
  publisher="International Speech Communication Association",
  address="Incheon",
  doi="10.21437/Interspeech.2022-10165",
  issn="1990-9772",
  url="https://www.isca-speech.org/archive/pdfs/interspeech_2022/stafylakis22_interspeech.pdf"
}