How Are Deepfake Videos Detected? An Initial User Study

Nyein Nyein Thaw; Thin July; Aye Nu Wai; Dion Hoe Lian Goh; Alton Y.K. Chua

doi:10.1007/978-3-030-78635-9_80

How Are Deepfake Videos Detected? An Initial User Study

Nyein Nyein Thaw, Thin July, Aye Nu Wai, Dion Hoe Lian Goh^*, Alton Y.K. Chua

^*Corresponding author for this work

Nanyang Technological University

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

8 Citations (Scopus)

Abstract

This paper describes an exploratory study of deepfake video detection from a user’s perspective. Through semi-structured interviews, participants were asked to identify real and deepfake videos, and explain how they arrived at their conclusions. From the interviews, two sets of features were derived. One was associated with correct deepfake identification while the other was associated with incorrect identification. Interestingly, the two sets had overlapping features suggesting the difficulties associated with deepfake identification. Further, the majority of participants could not correctly identify all their assigned videos.

Original language	English
Title of host publication	HCI International 2021 - Posters - 23rd HCI International Conference, HCII 2021, Proceedings
Editors	Constantine Stephanidis, Margherita Antona, Stavroula Ntoa
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	631-636
Number of pages	6
ISBN (Print)	9783030786342
DOIs	https://doi.org/10.1007/978-3-030-78635-9_80
Publication status	Published - 2021
Externally published	Yes
Event	23rd International Conference on Human-Computer Interaction, HCII 2021 - Virtual, Online Duration: Jul 24 2021 → Jul 29 2021

Publication series

Name	Communications in Computer and Information Science
Volume	1419
ISSN (Print)	1865-0929
ISSN (Electronic)	1865-0937

Conference

Conference	23rd International Conference on Human-Computer Interaction, HCII 2021
City	Virtual, Online
Period	7/24/21 → 7/29/21

Bibliographical note

Publisher Copyright:
© 2021, Springer Nature Switzerland AG.

ASJC Scopus Subject Areas

General Computer Science
General Mathematics

Keywords

Deepfake videos
Identification
Misidentification
Misinformation
User study

Access to Document

10.1007/978-3-030-78635-9_80

Cite this

Thaw, N. N., July, T., Wai, A. N., Goh, D. H. L., & Chua, A. Y. K. (2021). How Are Deepfake Videos Detected? An Initial User Study. In C. Stephanidis, M. Antona, & S. Ntoa (Eds.), HCI International 2021 - Posters - 23rd HCI International Conference, HCII 2021, Proceedings (pp. 631-636). (Communications in Computer and Information Science; Vol. 1419). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-030-78635-9_80

Thaw, Nyein Nyein ; July, Thin ; Wai, Aye Nu et al. / How Are Deepfake Videos Detected? An Initial User Study. HCI International 2021 - Posters - 23rd HCI International Conference, HCII 2021, Proceedings. editor / Constantine Stephanidis ; Margherita Antona ; Stavroula Ntoa. Springer Science and Business Media Deutschland GmbH, 2021. pp. 631-636 (Communications in Computer and Information Science).

@inproceedings{be228a3ee8174c08a480afb6e1f214c0,

title = "How Are Deepfake Videos Detected? An Initial User Study",

abstract = "This paper describes an exploratory study of deepfake video detection from a user{\textquoteright}s perspective. Through semi-structured interviews, participants were asked to identify real and deepfake videos, and explain how they arrived at their conclusions. From the interviews, two sets of features were derived. One was associated with correct deepfake identification while the other was associated with incorrect identification. Interestingly, the two sets had overlapping features suggesting the difficulties associated with deepfake identification. Further, the majority of participants could not correctly identify all their assigned videos.",

keywords = "Deepfake videos, Identification, Misidentification, Misinformation, User study",

author = "Thaw, {Nyein Nyein} and Thin July and Wai, {Aye Nu} and Goh, {Dion Hoe Lian} and Chua, {Alton Y.K.}",

note = "Publisher Copyright: {\textcopyright} 2021, Springer Nature Switzerland AG.; 23rd International Conference on Human-Computer Interaction, HCII 2021 ; Conference date: 24-07-2021 Through 29-07-2021",

year = "2021",

doi = "10.1007/978-3-030-78635-9_80",

language = "English",

isbn = "9783030786342",

series = "Communications in Computer and Information Science",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "631--636",

editor = "Constantine Stephanidis and Margherita Antona and Stavroula Ntoa",

booktitle = "HCI International 2021 - Posters - 23rd HCI International Conference, HCII 2021, Proceedings",

address = "Germany",

}

Thaw, NN, July, T, Wai, AN, Goh, DHL & Chua, AYK 2021, How Are Deepfake Videos Detected? An Initial User Study. in C Stephanidis, M Antona & S Ntoa (eds), HCI International 2021 - Posters - 23rd HCI International Conference, HCII 2021, Proceedings. Communications in Computer and Information Science, vol. 1419, Springer Science and Business Media Deutschland GmbH, pp. 631-636, 23rd International Conference on Human-Computer Interaction, HCII 2021, Virtual, Online, 7/24/21. https://doi.org/10.1007/978-3-030-78635-9_80

How Are Deepfake Videos Detected? An Initial User Study. / Thaw, Nyein Nyein; July, Thin; Wai, Aye Nu et al.
HCI International 2021 - Posters - 23rd HCI International Conference, HCII 2021, Proceedings. ed. / Constantine Stephanidis; Margherita Antona; Stavroula Ntoa. Springer Science and Business Media Deutschland GmbH, 2021. p. 631-636 (Communications in Computer and Information Science; Vol. 1419).

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

TY - GEN

T1 - How Are Deepfake Videos Detected? An Initial User Study

AU - Thaw, Nyein Nyein

AU - July, Thin

AU - Wai, Aye Nu

AU - Goh, Dion Hoe Lian

AU - Chua, Alton Y.K.

PY - 2021

Y1 - 2021

N2 - This paper describes an exploratory study of deepfake video detection from a user’s perspective. Through semi-structured interviews, participants were asked to identify real and deepfake videos, and explain how they arrived at their conclusions. From the interviews, two sets of features were derived. One was associated with correct deepfake identification while the other was associated with incorrect identification. Interestingly, the two sets had overlapping features suggesting the difficulties associated with deepfake identification. Further, the majority of participants could not correctly identify all their assigned videos.

AB - This paper describes an exploratory study of deepfake video detection from a user’s perspective. Through semi-structured interviews, participants were asked to identify real and deepfake videos, and explain how they arrived at their conclusions. From the interviews, two sets of features were derived. One was associated with correct deepfake identification while the other was associated with incorrect identification. Interestingly, the two sets had overlapping features suggesting the difficulties associated with deepfake identification. Further, the majority of participants could not correctly identify all their assigned videos.

KW - Deepfake videos

KW - Identification

KW - Misidentification

KW - Misinformation

KW - User study

UR - http://www.scopus.com/inward/record.url?scp=85112077147&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85112077147&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-78635-9_80

DO - 10.1007/978-3-030-78635-9_80

M3 - Conference contribution

AN - SCOPUS:85112077147

SN - 9783030786342

T3 - Communications in Computer and Information Science

SP - 631

EP - 636

BT - HCI International 2021 - Posters - 23rd HCI International Conference, HCII 2021, Proceedings

A2 - Stephanidis, Constantine

A2 - Antona, Margherita

A2 - Ntoa, Stavroula

PB - Springer Science and Business Media Deutschland GmbH

T2 - 23rd International Conference on Human-Computer Interaction, HCII 2021

Y2 - 24 July 2021 through 29 July 2021

ER -

Thaw NN, July T, Wai AN, Goh DHL , Chua AYK. How Are Deepfake Videos Detected? An Initial User Study. In Stephanidis C, Antona M, Ntoa S, editors, HCI International 2021 - Posters - 23rd HCI International Conference, HCII 2021, Proceedings. Springer Science and Business Media Deutschland GmbH. 2021. p. 631-636. (Communications in Computer and Information Science). doi: 10.1007/978-3-030-78635-9_80

How Are Deepfake Videos Detected? An Initial User Study

Abstract

Publication series

Conference

Bibliographical note

ASJC Scopus Subject Areas

Keywords

Access to Document

Other files and links

Fingerprint

Cite this