Publication Details

BUT System Description to VoxCeleb Speaker Recognition Challenge 2019

ZEINALI, H.; WANG, S.; SILNOVA, A.; MATĚJKA, P.; PLCHOT, O. BUT System Description to VoxCeleb Speaker Recognition Challenge 2019. Proceedings of The VoxCeleb Challange Workshop 2019. Graz: 2019. p. 1-4.
Czech title
VUT systému pro VoxCeleb soutěž automatického rozpoznávání řečníka 2019
Type
conference paper
Language
English
Authors
Zeinali Hossein, Ph.D. (DCGM)
Wang Shuai
Silnova Anna, M.Sc., Ph.D. (DCGM)
Matějka Pavel, Ing., Ph.D.
Plchot Oldřich, Ing., Ph.D. (DCGM)
URL
Keywords

VoxCeleb Speaker Recognition Challenge, Deep NeuralNetworks, ResNet, x-vector, PLDA, Cosine distance

Abstract

In this report, we describe the submission of Brno Universityof Technology (BUT) team to the VoxCeleb Speaker RecognitionChallenge (VoxSRC) 2019. We also provide a brief analysisof different systems on VoxCeleb-1 test sets. Submittedsystems for both Fixed and Open conditions are a fusion of4 Convolutional Neural Network (CNN) topologies. The firstand second networks have ResNet34 topology and use twodimensionalCNNs. The last two networks are one-dimensionalCNN and are based on the x-vector extraction topology. Someof the networks are fine-tuned using additive margin angularsoftmax. Kaldi FBanks and Kaldi PLPs were used as features.The difference between Fixed and Open systems lies in the usedtraining data and fusion strategy. The best systems for Fixedand Open conditions achieved 1.42 % and 1.26 % ERR on thechallenge evaluation set respectively.

Published
2019
Pages
1–4
Proceedings
Proceedings of The VoxCeleb Challange Workshop 2019
Place
Graz
BibTeX
@inproceedings{BUT168476,
  author="Hossein {Zeinali} and Shuai {Wang} and Anna {Silnova} and Pavel {Matějka} and Oldřich {Plchot}",
  title="BUT System Description to VoxCeleb Speaker Recognition Challenge 2019",
  booktitle="Proceedings of The VoxCeleb Challange Workshop 2019",
  year="2019",
  pages="1--4",
  address="Graz",
  url="https://arxiv.org/abs/1910.12592"
}
Files
Back to top