Adversarial attacks on skeleton-based sign language recognition

Yufeng Li; Meng Han; Jiahui Yu; Changting Lin; Zhaojie Ju

doi:10.1007/978-981-99-6483-3_4

Adversarial attacks on skeleton-based sign language recognition

Yufeng Li
, Meng Han^*
, Jiahui Yu
, Changting Lin
, Zhaojie Ju

^*Corresponding author for this work

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

61 Downloads (Pure)

Abstract

Despite the impressive performance achieved by sign language recognition systems based on skeleton information, our research has uncovered their vulnerability to malicious attacks. In response to this challenge, we present an adversarial attack specifically designed to sign language recognition models that rely on extracted human skeleton data as features. Our attack aims to assess the robustness and sensitivity of these models, and we propose adversarial training techniques to enhance their resilience. Moreover, we conduct transfer experiments using the generated adversarial samples to demonstrate the transferability of these adversarial examples across different models. Additionally, by conducting experiments on the sensitivity of sign language recognition models, we identify the optimal experimental parameter settings for achieving the most effective attacks. This research significantly contributes to future investigations into the security of sign language recognition.

Original language	English
Title of host publication	Intelligent Robotics and Applications - 16th International Conference, ICIRA 2023, Proceedings
Editors	Huayong Yang, Jun Zou, Geng Yang, Xiaoping Ouyang, Honghai Liu, Zhiyong Wang, Zhouping Yin, Lianqing Liu
Publisher	Springer
Pages	33-43
Number of pages	11
ISBN (Electronic)	9789819964833
ISBN (Print)	9789819964826
DOIs	https://doi.org/10.1007/978-981-99-6483-3_4 https://doi.org/10.1007/978-981-99-6483-3
Publication status	Published - 21 Oct 2023
Event	16th International Conference on Intelligent Robotics and Applications, ICIRA 2023 - Hangzhou, China Duration: 5 Jul 2023 → 7 Jul 2023

Publication series

Name	Lecture Notes in Computer Science
Publisher	Springer
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Name	Lecture Notes in Artificial Intelligence
Publisher	Springer
Volume	14267
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	16th International Conference on Intelligent Robotics and Applications, ICIRA 2023
Country/Territory	China
City	Hangzhou
Period	5/07/23 → 7/07/23

Keywords

adversarial attacks
robustness
sign language recognition

Access to Document

Li_et_al_2023_AAM
This version of the article has been accepted for publication, after peer review (when applicable) and is subject to Springer Nature’s AM terms of use, but is not the Version of Record and does not reflect post-acceptance improvements, or any corrections. The Version of Record is available online at: http://dx.doi.org/10.1007/978-981-99-6483-3_4.
Accepted author manuscript (Post-print), 387 KB

Cite this

Li, Y., Han, M., Yu, J., Lin, C., & Ju, Z. (2023). Adversarial attacks on skeleton-based sign language recognition. In H. Yang, J. Zou, G. Yang, X. Ouyang, H. Liu, Z. Wang, Z. Yin, & L. Liu (Eds.), Intelligent Robotics and Applications - 16th International Conference, ICIRA 2023, Proceedings (pp. 33-43). (Lecture Notes in Computer Science), (Lecture Notes in Artificial Intelligence; Vol. 14267). Springer. https://doi.org/10.1007/978-981-99-6483-3_4, https://doi.org/10.1007/978-981-99-6483-3

Li, Yufeng ; Han, Meng ; Yu, Jiahui et al. / Adversarial attacks on skeleton-based sign language recognition. Intelligent Robotics and Applications - 16th International Conference, ICIRA 2023, Proceedings. editor / Huayong Yang ; Jun Zou ; Geng Yang ; Xiaoping Ouyang ; Honghai Liu ; Zhiyong Wang ; Zhouping Yin ; Lianqing Liu. Springer, 2023. pp. 33-43 (Lecture Notes in Computer Science). (Lecture Notes in Artificial Intelligence).

@inproceedings{9e5eb02210ed475d96709ae907f4a795,

title = "Adversarial attacks on skeleton-based sign language recognition",

abstract = "Despite the impressive performance achieved by sign language recognition systems based on skeleton information, our research has uncovered their vulnerability to malicious attacks. In response to this challenge, we present an adversarial attack specifically designed to sign language recognition models that rely on extracted human skeleton data as features. Our attack aims to assess the robustness and sensitivity of these models, and we propose adversarial training techniques to enhance their resilience. Moreover, we conduct transfer experiments using the generated adversarial samples to demonstrate the transferability of these adversarial examples across different models. Additionally, by conducting experiments on the sensitivity of sign language recognition models, we identify the optimal experimental parameter settings for achieving the most effective attacks. This research significantly contributes to future investigations into the security of sign language recognition.",

keywords = "adversarial attacks, robustness, sign language recognition",

author = "Yufeng Li and Meng Han and Jiahui Yu and Changting Lin and Zhaojie Ju",

year = "2023",

month = oct,

day = "21",

doi = "10.1007/978-981-99-6483-3\_4",

language = "English",

isbn = "9789819964826",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "33--43",

editor = "Huayong Yang and Jun Zou and Geng Yang and Xiaoping Ouyang and Honghai Liu and Zhiyong Wang and Zhouping Yin and Lianqing Liu",

booktitle = "Intelligent Robotics and Applications - 16th International Conference, ICIRA 2023, Proceedings",

address = "Netherlands",

note = "16th International Conference on Intelligent Robotics and Applications, ICIRA 2023 ; Conference date: 05-07-2023 Through 07-07-2023",

}

Li, Y, Han, M, Yu, J, Lin, C & Ju, Z 2023, Adversarial attacks on skeleton-based sign language recognition. in H Yang, J Zou, G Yang, X Ouyang, H Liu, Z Wang, Z Yin & L Liu (eds), Intelligent Robotics and Applications - 16th International Conference, ICIRA 2023, Proceedings. Lecture Notes in Computer Science, Lecture Notes in Artificial Intelligence, vol. 14267, Springer, pp. 33-43, 16th International Conference on Intelligent Robotics and Applications, ICIRA 2023, Hangzhou, China, 5/07/23. https://doi.org/10.1007/978-981-99-6483-3_4, https://doi.org/10.1007/978-981-99-6483-3

Adversarial attacks on skeleton-based sign language recognition. / Li, Yufeng; Han, Meng; Yu, Jiahui et al.
Intelligent Robotics and Applications - 16th International Conference, ICIRA 2023, Proceedings. ed. / Huayong Yang; Jun Zou; Geng Yang; Xiaoping Ouyang; Honghai Liu; Zhiyong Wang; Zhouping Yin; Lianqing Liu. Springer, 2023. p. 33-43 (Lecture Notes in Computer Science), (Lecture Notes in Artificial Intelligence; Vol. 14267).

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

TY - GEN

T1 - Adversarial attacks on skeleton-based sign language recognition

AU - Li, Yufeng

AU - Han, Meng

AU - Yu, Jiahui

AU - Lin, Changting

AU - Ju, Zhaojie

PY - 2023/10/21

Y1 - 2023/10/21

N2 - Despite the impressive performance achieved by sign language recognition systems based on skeleton information, our research has uncovered their vulnerability to malicious attacks. In response to this challenge, we present an adversarial attack specifically designed to sign language recognition models that rely on extracted human skeleton data as features. Our attack aims to assess the robustness and sensitivity of these models, and we propose adversarial training techniques to enhance their resilience. Moreover, we conduct transfer experiments using the generated adversarial samples to demonstrate the transferability of these adversarial examples across different models. Additionally, by conducting experiments on the sensitivity of sign language recognition models, we identify the optimal experimental parameter settings for achieving the most effective attacks. This research significantly contributes to future investigations into the security of sign language recognition.

AB - Despite the impressive performance achieved by sign language recognition systems based on skeleton information, our research has uncovered their vulnerability to malicious attacks. In response to this challenge, we present an adversarial attack specifically designed to sign language recognition models that rely on extracted human skeleton data as features. Our attack aims to assess the robustness and sensitivity of these models, and we propose adversarial training techniques to enhance their resilience. Moreover, we conduct transfer experiments using the generated adversarial samples to demonstrate the transferability of these adversarial examples across different models. Additionally, by conducting experiments on the sensitivity of sign language recognition models, we identify the optimal experimental parameter settings for achieving the most effective attacks. This research significantly contributes to future investigations into the security of sign language recognition.

KW - adversarial attacks

KW - robustness

KW - sign language recognition

UR - https://www.scopus.com/pages/publications/85175956607

U2 - 10.1007/978-981-99-6483-3_4

DO - 10.1007/978-981-99-6483-3_4

M3 - Conference contribution

AN - SCOPUS:85175956607

SN - 9789819964826

T3 - Lecture Notes in Computer Science

SP - 33

EP - 43

BT - Intelligent Robotics and Applications - 16th International Conference, ICIRA 2023, Proceedings

A2 - Yang, Huayong

A2 - Zou, Jun

A2 - Yang, Geng

A2 - Ouyang, Xiaoping

A2 - Liu, Honghai

A2 - Wang, Zhiyong

A2 - Yin, Zhouping

A2 - Liu, Lianqing

PB - Springer

T2 - 16th International Conference on Intelligent Robotics and Applications, ICIRA 2023

Y2 - 5 July 2023 through 7 July 2023

ER -

Li Y, Han M, Yu J, Lin C, Ju Z. Adversarial attacks on skeleton-based sign language recognition. In Yang H, Zou J, Yang G, Ouyang X, Liu H, Wang Z, Yin Z, Liu L, editors, Intelligent Robotics and Applications - 16th International Conference, ICIRA 2023, Proceedings. Springer. 2023. p. 33-43. (Lecture Notes in Computer Science). (Lecture Notes in Artificial Intelligence). doi: 10.1007/978-981-99-6483-3_4, 10.1007/978-981-99-6483-3

Adversarial attacks on skeleton-based sign language recognition

Abstract

Publication series

Conference

Keywords

Access to Document

Fingerprint

Cite this