Loading...
Loading...

Go to the content (press return)

Auto-encoding nearest neighbor i-vectors for speaker verification

Author
Khan, U.; India, M.; Hernando, J.
Type of activity
Presentation of work at congresses
Name of edition
20th Annual Conference of the International Speech Communication Association
Date of publication
2019
Presentation's date
2019-09-16
Book of congress proceedings
Interspeech 2019: the 20th Annual Conference of the International Speech Communication Association: 15-19 September 2019: Graz, Austria
First page
4060
Last page
4064
Publisher
International Speech Communication Association (ISCA)
DOI
10.21437/Interspeech.2019-1444
Project funding
Deep learning technologies for speech and audio processing
Repository
http://hdl.handle.net/2117/178617 Open in new window
URL
https://www.isca-speech.org/archive/Interspeech_2019/pdfs/1444.pdf Open in new window
Abstract
In the last years, i-vectors followed by cosine or PLDA scoringtechniques were the state-of-the-art approach in speaker veri-fication. PLDA requires labeled background data, and thereexists a significant performance gap between the two scoringtechniques. In this work, we propose to reduce this gap by us-ing an autoencoder to transform i-vector into a new speaker vec-tor representation, which will be referred to as ae-vector. Theautoencoder will be trained to reconstruct neighbor i-vectors in-ste...
Citation
Khan, U.; India, M.; Hernando, J. Auto-encoding nearest neighbor i-vectors for speaker verification. A: Annual Conference of the International Speech Communication Association. "Interspeech 2019: the 20th Annual Conference of the International Speech Communication Association: 15-19 September 2019: Graz, Austria". Baixas: International Speech Communication Association (ISCA), 2019, p. 4060-4064.
Keywords
Autoencoders, Deep learning, Speaker verification, i-vectors
Group of research
IDEAI-UPC - Intelligent Data Science and Artificial Intelligence Research Center
TALP - Centre for Language and Speech Technologies and Applications
VEU - Speech Processing Group

Participants

Attachments