Unsupervised learning in cross-corpus acoustic emotion recognition

Zixing Zhang; Felix Weninger; Martin Wöllmer; Björn Schuller

doi:10.1109/ASRU.2011.6163986

Unsupervised learning in cross-corpus acoustic emotion recognition

Zixing Zhang, Felix Weninger, Martin Wöllmer, Björn Schuller

Chair of Health Informatics

Technical University of Munich

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

102 Scopus citations

Abstract

One of the ever-present bottlenecks in Automatic Emotion Recognition is data sparseness. We therefore investigate the suitability of unsupervised learning in cross-corpus acoustic emotion recognition through a large-scale study with six commonly used databases, including acted and natural emotion speech, and covering a variety of application scenarios and acoustic conditions. We show that adding unlabeled emotional speech to agglomerated multi-corpus training sets can enhance recognition performance even in a challenging cross-corpus setting; furthermore, we show that the expected gain by adding unlabeled data on average is approximately half the one achieved by additional manually labeled data in leave-one-corpus-out validation.

Original language	English
Title of host publication	2011 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2011, Proceedings
Pages	523-528
Number of pages	6
DOIs	https://doi.org/10.1109/ASRU.2011.6163986
State	Published - 2011
Event	2011 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2011 - Waikoloa, HI, United States Duration: 11 Dec 2011 → 15 Dec 2011

Publication series

Name	2011 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2011, Proceedings

Conference

Conference	2011 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2011
Country/Territory	United States
City	Waikoloa, HI
Period	11/12/11 → 15/12/11

Keywords

speech emotion recognition
unsupervised learning

Access to Document

10.1109/ASRU.2011.6163986

Cite this

Zhang, Z., Weninger, F., Wöllmer, M., & Schuller, B. (2011). Unsupervised learning in cross-corpus acoustic emotion recognition. In 2011 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2011, Proceedings (pp. 523-528). Article 6163986 (2011 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2011, Proceedings). https://doi.org/10.1109/ASRU.2011.6163986

@inproceedings{fa26d6b890e2415b8ee4fbec7a085a8c,

title = "Unsupervised learning in cross-corpus acoustic emotion recognition",

abstract = "One of the ever-present bottlenecks in Automatic Emotion Recognition is data sparseness. We therefore investigate the suitability of unsupervised learning in cross-corpus acoustic emotion recognition through a large-scale study with six commonly used databases, including acted and natural emotion speech, and covering a variety of application scenarios and acoustic conditions. We show that adding unlabeled emotional speech to agglomerated multi-corpus training sets can enhance recognition performance even in a challenging cross-corpus setting; furthermore, we show that the expected gain by adding unlabeled data on average is approximately half the one achieved by additional manually labeled data in leave-one-corpus-out validation.",

keywords = "speech emotion recognition, unsupervised learning",

author = "Zixing Zhang and Felix Weninger and Martin W{\"o}llmer and Bj{\"o}rn Schuller",

year = "2011",

doi = "10.1109/ASRU.2011.6163986",

language = "English",

isbn = "9781467303675",

series = "2011 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2011, Proceedings",

pages = "523--528",

booktitle = "2011 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2011, Proceedings",

note = "2011 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2011 ; Conference date: 11-12-2011 Through 15-12-2011",

}

Zhang, Z, Weninger, F, Wöllmer, M & Schuller, B 2011, Unsupervised learning in cross-corpus acoustic emotion recognition. in 2011 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2011, Proceedings., 6163986, 2011 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2011, Proceedings, pp. 523-528, 2011 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2011, Waikoloa, HI, United States, 11/12/11. https://doi.org/10.1109/ASRU.2011.6163986

Unsupervised learning in cross-corpus acoustic emotion recognition. / Zhang, Zixing; Weninger, Felix; Wöllmer, Martin et al.
2011 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2011, Proceedings. 2011. p. 523-528 6163986 (2011 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2011, Proceedings).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Unsupervised learning in cross-corpus acoustic emotion recognition

AU - Zhang, Zixing

AU - Weninger, Felix

AU - Wöllmer, Martin

AU - Schuller, Björn

PY - 2011

Y1 - 2011

N2 - One of the ever-present bottlenecks in Automatic Emotion Recognition is data sparseness. We therefore investigate the suitability of unsupervised learning in cross-corpus acoustic emotion recognition through a large-scale study with six commonly used databases, including acted and natural emotion speech, and covering a variety of application scenarios and acoustic conditions. We show that adding unlabeled emotional speech to agglomerated multi-corpus training sets can enhance recognition performance even in a challenging cross-corpus setting; furthermore, we show that the expected gain by adding unlabeled data on average is approximately half the one achieved by additional manually labeled data in leave-one-corpus-out validation.

AB - One of the ever-present bottlenecks in Automatic Emotion Recognition is data sparseness. We therefore investigate the suitability of unsupervised learning in cross-corpus acoustic emotion recognition through a large-scale study with six commonly used databases, including acted and natural emotion speech, and covering a variety of application scenarios and acoustic conditions. We show that adding unlabeled emotional speech to agglomerated multi-corpus training sets can enhance recognition performance even in a challenging cross-corpus setting; furthermore, we show that the expected gain by adding unlabeled data on average is approximately half the one achieved by additional manually labeled data in leave-one-corpus-out validation.

KW - speech emotion recognition

KW - unsupervised learning

UR - http://www.scopus.com/inward/record.url?scp=84858985413&partnerID=8YFLogxK

U2 - 10.1109/ASRU.2011.6163986

DO - 10.1109/ASRU.2011.6163986

M3 - Conference contribution

AN - SCOPUS:84858985413

SN - 9781467303675

T3 - 2011 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2011, Proceedings

SP - 523

EP - 528

BT - 2011 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2011, Proceedings

T2 - 2011 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2011

Y2 - 11 December 2011 through 15 December 2011

ER -

Unsupervised learning in cross-corpus acoustic emotion recognition

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this