Publication Details

Listen only to me! How well can target speech extraction handle false alarms?

DELCROIX Marc, KINOSHITA Keisuke, OCHIAI Tsubasa, ŽMOLÍKOVÁ Kateřina, SATO Hiroshi and NAKATANI Tomohiro. Listen only to me! How well can target speech extraction handle false alarms?. In: Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH. Incheon: International Speech Communication Association, 2022, pp. 216-220. ISSN 1990-9772. Available from: https://www.isca-speech.org/archive/pdfs/interspeech_2022/delcroix22_interspeech.pdf
Czech title
Poslouchej jen mě! Jak dobře zvládá extrakce řeči cílového mluvčího falešné poplachy?
Type
conference paper
Language
english
Authors
Delcroix Marc (NTT)
Kinoshita Keisuke (NTT)
Ochiai Tsubasa (NTT)
Žmolíková Kateřina, Ing., Ph.D. (DCGM FIT BUT)
Sato Hiroshi (NTT)
Nakatani Tomohiro (NTT)
URL
Keywords

Speech enhancement, Target speech extraction, Inactive speaker

Abstract

Target speech extraction (TSE) extracts the speech of a target speaker in a mixture given auxiliary clues characterizing the speaker, such as an enrollment utterance. TSE addresses thus the challenging problem of simultaneously performing separation and speaker identification. There has been much progress in extraction performance following the recent development of neural networks for speech enhancement and separation. Most studies have focused on processing mixtures where the target speaker is actively speaking. However, the target speaker is sometimes silent in practice, i.e., inactive speaker (IS). A typical TSE system will tend to output a signal in IS cases, causing false alarms. This is a severe problem for the practical deployment of TSE systems. This paper aims at understanding better how well TSE systems can handle IS cases. We consider two approaches to deal with IS, (1) training a system to directly output zero signals or (2) detecting IS with an extra speaker verification module. We perform an extensive experimental comparison of these schemes in terms of extraction performance and IS detection using the LibriMix dataset and reveal their pros and cons.

Published
2022
Pages
216-220
Journal
Proceedings of Interspeech - on-line, no. 9, ISSN 1990-9772
Proceedings
Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH
Conference
Interspeech Conference, Incheon, KR
Publisher
International Speech Communication Association
Place
Incheon, KR
DOI
10.21437/Interspeech.2022-11252
EID Scopus
2-s2.0-85140085197
BibTeX 
@INPROCEEDINGS{FITPUB12855,
   author = "Marc Delcroix and Keisuke Kinoshita and Tsubasa Ochiai and Kate\v{r}ina \v{Z}mol\'{i}kov\'{a} and Hiroshi Sato and Tomohiro Nakatani",
   title = "Listen only to me! How well can target speech extraction handle false alarms?",
   pages = "216--220",
   booktitle = "Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH",
   journal = "Proceedings of Interspeech - on-line",
   number = 9,
   year = 2022,
   location = "Incheon, KR",
   publisher = "International Speech Communication Association",
   ISSN = "1990-9772",
   doi = "10.21437/Interspeech.2022-11252",
   language = "english",
   url = "https://www.fit.vut.cz/research/publication/12855"
}
Back to top