Speaker Diarization Using Deep Recurrent Convolutional Neural Networks for Speaker Embeddings

Pawel Cyrta , Tomasz Trzciński , Wojciech Stokowiec


In this paper we propose a new method of speaker diarization that employs a deep learning architecture to learn speaker embeddings. In contrast to the traditional approaches that build their speaker embeddings using manually hand-crafted spectral features, we propose to train for this purpose a recurrent convolutional neural network applied directly on magnitude spectrograms. To compare our approach with the state of the art, we collect and release for the public an additional dataset of over 6 h of fully annotated broadcast material. The results of our evaluation on the new dataset and three other benchmark datasets show that our proposed method significantly outperforms the competitors and reduces diarization error rate by a large margin of over 30% with respect to the baseline.
Author Pawel Cyrta
Pawel Cyrta,,
, Tomasz Trzciński II
Tomasz Trzciński,,
- The Institute of Computer Science
, Wojciech Stokowiec
Wojciech Stokowiec,,
Publication size in sheets0.5
Book Borzemski Leszek, Świątek Jerzy, Wilimowska Zofia (eds.): Information Systems Architecture and Technology: Proceedings of 38th International Conference on Information Systems Architecture and Technology – ISAT 2017. Part I, Advances in Intelligent Systems and Computing, vol. 655, 2018, Springer International Publishing, ISBN 978-3-319-67219-9, [978-3-319-67220-5], 358 p., DOI:10.1007/978-3-319-67220-5
Keywords in EnglishSpeaker diarization, Speaker embeddings, Speaker clustering, Deep neural network, Recursive convolutional neural networks, Convolutional neural networks 
URL https://link.springer.com/chapter/10.1007/978-3-319-67220-5_10
Languageen angielski
Score (nominal)0
Citation count*0
Share Share

Get link to the record

* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.