Automatic Detection of Backchannels in Russian Dialogue Speech

Pavel  Kholiavin; Anna  Mamushina; Daniil  Kocharov; Tatiana Kachkovskaia

doi:https://doi.org/10.1007/978-3-030-60276-5_21

Automatic Detection of Backchannels in Russian Dialogue Speech

Pavel Kholiavin, Anna Mamushina, Daniil Kocharov, Tatiana Kachkovskaia

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

Abstract

This paper deals with acoustic properties of backchannels – those turns within a dialogue which do not convey information but signify that the speaker is listening to his/her interlocutor (uh-huh, hm etc.). The research is based on a Russian corpus of dialogue speech, SibLing, a part of which (339 min of speech) was manually segmented into backchannels and non-backchannels. Then, a number of acoustic parameters was calculated: duration, intensity, fundamental frequency, and pause duration. Our data have shown that in Russian speech backchannels are shorter and have lower loudness and pitch than non-backchannels. After that, two classifiers were tested: CART and SVM. The highest efficiency was achieved using SVM (F 1 = 0.651) and the following feature set: duration, maximum fundamental frequency, melodic slope. The most valuable feature was duration.

Original language	English
Title of host publication	Speech and Computer
Subtitle of host publication	22nd International Conference, SPECOM 2020, St. Petersburg, Russia, October 7–9, 2020, Proceedings
Editors	Alexey Karpov, Rodmonga Potapova
Place of Publication	Cham
Publisher	Springer Nature
Pages	204-213
ISBN (Electronic)	978-3-030-60276-5
ISBN (Print)	978-3-030-60275-8
DOIs	https://doi.org/10.1007/978-3-030-60276-5_21
State	Published - 2020
Event	22nd International Conference on Speech and Computer - St. Petersburg, Russia => Online, St. Petersburg, Russian Federation Duration: 7 Oct 2020 → 9 Oct 2020 http://specom.nw.ru/2020/program/SPECOM-ICR2020-Conference-Program-06102020.pdf

Publication series

Name	Lecture Notes in Computer Science
Volume	12335
ISSN (Print)	0302-9743

Conference

Conference	22nd International Conference on Speech and Computer
Abbreviated title	SPECOM 2020
Country/Territory	Russian Federation
City	St. Petersburg
Period	7/10/20 → 9/10/20
Internet address	http://specom.nw.ru/2020/program/SPECOM-ICR2020-Conference-Program-06102020.pdf

Access to Document

https://doi.org/10.1007/978-3-030-60276-5_21

Cite this

Kholiavin, P., Mamushina, A., Kocharov, D., & Kachkovskaia, T. (2020). Automatic Detection of Backchannels in Russian Dialogue Speech. In A. Karpov, & R. Potapova (Eds.), Speech and Computer: 22nd International Conference, SPECOM 2020, St. Petersburg, Russia, October 7–9, 2020, Proceedings (pp. 204-213). (Lecture Notes in Computer Science; Vol. 12335). Springer Nature. https://doi.org/10.1007/978-3-030-60276-5_21

Kholiavin, Pavel ; Mamushina, Anna ; Kocharov, Daniil ; Kachkovskaia, Tatiana. / Automatic Detection of Backchannels in Russian Dialogue Speech. Speech and Computer: 22nd International Conference, SPECOM 2020, St. Petersburg, Russia, October 7–9, 2020, Proceedings. editor / Alexey Karpov ; Rodmonga Potapova. Cham : Springer Nature, 2020. pp. 204-213 (Lecture Notes in Computer Science).

@inproceedings{117bfb64ffb54bfa84844285ef486d0b,

title = "Automatic Detection of Backchannels in Russian Dialogue Speech",

abstract = "This paper deals with acoustic properties of backchannels – those turns within a dialogue which do not convey information but signify that the speaker is listening to his/her interlocutor (uh-huh, hm etc.). The research is based on a Russian corpus of dialogue speech, SibLing, a part of which (339 min of speech) was manually segmented into backchannels and non-backchannels. Then, a number of acoustic parameters was calculated: duration, intensity, fundamental frequency, and pause duration. Our data have shown that in Russian speech backchannels are shorter and have lower loudness and pitch than non-backchannels. After that, two classifiers were tested: CART and SVM. The highest efficiency was achieved using SVM (F 1 = 0.651) and the following feature set: duration, maximum fundamental frequency, melodic slope. The most valuable feature was duration.",

keywords = "dialogue speech, backchannel, turn-taking, speech acoustics, Russian",

author = "Pavel Kholiavin and Anna Mamushina and Daniil Kocharov and Tatiana Kachkovskaia",

note = "Kholiavin P., Mamushina A., Kocharov D., Kachkovskaia T. (2020) Automatic Detection of Backchannels in Russian Dialogue Speech. In: Karpov A., Potapova R. (eds) Speech and Computer. SPECOM 2020. Lecture Notes in Computer Science, vol 12335. Springer, Cham. https://doi.org/10.1007/978-3-030-60276-5_21; 22nd International Conference on Speech and Computer, SPECOM 2020 ; Conference date: 07-10-2020 Through 09-10-2020",

year = "2020",

doi = "https://doi.org/10.1007/978-3-030-60276-5_21",

language = "English",

isbn = "978-3-030-60275-8",

series = "Lecture Notes in Computer Science",

publisher = "Springer Nature",

pages = "204--213",

editor = "Karpov, {Alexey } and Potapova, {Rodmonga }",

booktitle = "Speech and Computer",

address = "Germany",

url = "http://specom.nw.ru/2020/program/SPECOM-ICR2020-Conference-Program-06102020.pdf",

}

Kholiavin, P, Mamushina, A, Kocharov, D & Kachkovskaia, T 2020, Automatic Detection of Backchannels in Russian Dialogue Speech. in A Karpov & R Potapova (eds), Speech and Computer: 22nd International Conference, SPECOM 2020, St. Petersburg, Russia, October 7–9, 2020, Proceedings. Lecture Notes in Computer Science, vol. 12335, Springer Nature, Cham, pp. 204-213, 22nd International Conference on Speech and Computer, St. Petersburg, Russian Federation, 7/10/20. https://doi.org/10.1007/978-3-030-60276-5_21

Automatic Detection of Backchannels in Russian Dialogue Speech. / Kholiavin, Pavel ; Mamushina, Anna ; Kocharov, Daniil ; Kachkovskaia, Tatiana.

Speech and Computer: 22nd International Conference, SPECOM 2020, St. Petersburg, Russia, October 7–9, 2020, Proceedings. ed. / Alexey Karpov; Rodmonga Potapova. Cham : Springer Nature, 2020. p. 204-213 (Lecture Notes in Computer Science; Vol. 12335).

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

TY - GEN

T1 - Automatic Detection of Backchannels in Russian Dialogue Speech

AU - Kholiavin, Pavel

AU - Mamushina, Anna

AU - Kocharov, Daniil

AU - Kachkovskaia, Tatiana

N1 - Kholiavin P., Mamushina A., Kocharov D., Kachkovskaia T. (2020) Automatic Detection of Backchannels in Russian Dialogue Speech. In: Karpov A., Potapova R. (eds) Speech and Computer. SPECOM 2020. Lecture Notes in Computer Science, vol 12335. Springer, Cham. https://doi.org/10.1007/978-3-030-60276-5_21

PY - 2020

Y1 - 2020

N2 - This paper deals with acoustic properties of backchannels – those turns within a dialogue which do not convey information but signify that the speaker is listening to his/her interlocutor (uh-huh, hm etc.). The research is based on a Russian corpus of dialogue speech, SibLing, a part of which (339 min of speech) was manually segmented into backchannels and non-backchannels. Then, a number of acoustic parameters was calculated: duration, intensity, fundamental frequency, and pause duration. Our data have shown that in Russian speech backchannels are shorter and have lower loudness and pitch than non-backchannels. After that, two classifiers were tested: CART and SVM. The highest efficiency was achieved using SVM (F 1 = 0.651) and the following feature set: duration, maximum fundamental frequency, melodic slope. The most valuable feature was duration.

AB - This paper deals with acoustic properties of backchannels – those turns within a dialogue which do not convey information but signify that the speaker is listening to his/her interlocutor (uh-huh, hm etc.). The research is based on a Russian corpus of dialogue speech, SibLing, a part of which (339 min of speech) was manually segmented into backchannels and non-backchannels. Then, a number of acoustic parameters was calculated: duration, intensity, fundamental frequency, and pause duration. Our data have shown that in Russian speech backchannels are shorter and have lower loudness and pitch than non-backchannels. After that, two classifiers were tested: CART and SVM. The highest efficiency was achieved using SVM (F 1 = 0.651) and the following feature set: duration, maximum fundamental frequency, melodic slope. The most valuable feature was duration.

KW - dialogue speech

KW - backchannel

KW - turn-taking

KW - speech acoustics

KW - Russian

UR - https://link.springer.com/book/10.1007/978-3-030-60276-5

U2 - https://doi.org/10.1007/978-3-030-60276-5_21

DO - https://doi.org/10.1007/978-3-030-60276-5_21

M3 - Conference contribution

SN - 978-3-030-60275-8

T3 - Lecture Notes in Computer Science

SP - 204

EP - 213

BT - Speech and Computer

A2 - Karpov, Alexey

A2 - Potapova, Rodmonga

PB - Springer Nature

CY - Cham

T2 - 22nd International Conference on Speech and Computer

Y2 - 7 October 2020 through 9 October 2020

ER -

Kholiavin P, Mamushina A, Kocharov D , Kachkovskaia T. Automatic Detection of Backchannels in Russian Dialogue Speech. In Karpov A, Potapova R, editors, Speech and Computer: 22nd International Conference, SPECOM 2020, St. Petersburg, Russia, October 7–9, 2020, Proceedings. Cham: Springer Nature. 2020. p. 204-213. (Lecture Notes in Computer Science). https://doi.org/10.1007/978-3-030-60276-5_21

Automatic Detection of Backchannels in Russian Dialogue Speech

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this