De-Identifying Facial Images Using Projections on Hyperspheres

P. Chriskos; O. Zoidi; A. Tefas; Ioannis Pitas

doi:10.1109/FG.2015.7285020

De-Identifying Facial Images Using Projections on Hyperspheres

P. Chriskos, O. Zoidi, A. Tefas, Ioannis Pitas

Department of Electrical & Electronic Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference Contribution (Conference Proceeding)

230 Downloads (Pure)

Abstract

A major issue that arises from mass visual media distribution in modern video sharing, social media and cloud services, is the issue of privacy. Malicious users can use these services to track the actions of certain individuals and/or groups
thus violating their privacy. As a result the need to hinder automatic facial image identification in images and videos arises. In this paper we propose a method for de-identifying facial images. Contrary to most de-identification methods, this
method manipulates facial images so that humans can still recognize the individual or individuals in an image or video frame, but at the same time common automatic identification algorithms fail to do so. This is achieved by projecting the facial images on a hypersphere. From the conducted experiments
it can be verified that this method is effective in reducing the classification accuracy under 10%. Furthermore, in the resulting images the subject can be identified by human viewers.

Original language	English
Title of host publication	2015 11th IEEE International Conference and Workshops on Automatic Face and Gesture Recognition (FG 2015)
Subtitle of host publication	Proceedings of a meeting held 4-8 May 2015, Ljubljana, Slovenia
Publisher	Institute of Electrical and Electronics Engineers (IEEE)
Pages	1-6
Number of pages	6
Volume	4
ISBN (Electronic)	9781479960262
ISBN (Print)	9781479960279
DOIs	https://doi.org/10.1109/FG.2015.7285020
Publication status	Published - Aug 2015
Event	Automatic Face and Gesture Recognition (FG), 11th IEEE International Conference and Workshops - Ljubljana, Slovenia Duration: 4 May 2015 → 8 May 2015

Conference

Conference	Automatic Face and Gesture Recognition (FG), 11th IEEE International Conference and Workshops
Country	Slovenia
City	Ljubljana
Period	4/05/15 → 8/05/15

Access to Document

10.1109/FG.2015.7285020

Ioannis Pitas - De-identifying facial images using projections on hyperspheres - 2015
This is the author accepted manuscript (AAM). The final published version (version of record) is available online via IEEE at http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=7285020. Please refer to any applicable terms of use of the publisher.
Accepted author manuscript, 1.24 MB

Cite this

Chriskos, P., Zoidi, O., Tefas, A., & Pitas, I. (2015). De-Identifying Facial Images Using Projections on Hyperspheres. In 2015 11th IEEE International Conference and Workshops on Automatic Face and Gesture Recognition (FG 2015): Proceedings of a meeting held 4-8 May 2015, Ljubljana, Slovenia (Vol. 4, pp. 1-6). Institute of Electrical and Electronics Engineers (IEEE). https://doi.org/10.1109/FG.2015.7285020

Chriskos, P. ; Zoidi, O. ; Tefas, A. ; Pitas, Ioannis. / De-Identifying Facial Images Using Projections on Hyperspheres. 2015 11th IEEE International Conference and Workshops on Automatic Face and Gesture Recognition (FG 2015): Proceedings of a meeting held 4-8 May 2015, Ljubljana, Slovenia. Vol. 4 Institute of Electrical and Electronics Engineers (IEEE), 2015. pp. 1-6

@inproceedings{dd9f70a8b84f417dbb2ff21a74f2124d,

title = "De-Identifying Facial Images Using Projections on Hyperspheres",

abstract = "A major issue that arises from mass visual media distribution in modern video sharing, social media and cloud services, is the issue of privacy. Malicious users can use these services to track the actions of certain individuals and/or groupsthus violating their privacy. As a result the need to hinder automatic facial image identification in images and videos arises. In this paper we propose a method for de-identifying facial images. Contrary to most de-identification methods, thismethod manipulates facial images so that humans can still recognize the individual or individuals in an image or video frame, but at the same time common automatic identification algorithms fail to do so. This is achieved by projecting the facial images on a hypersphere. From the conducted experimentsit can be verified that this method is effective in reducing the classification accuracy under 10%. Furthermore, in the resulting images the subject can be identified by human viewers.",

author = "P. Chriskos and O. Zoidi and A. Tefas and Ioannis Pitas",

year = "2015",

month = aug,

doi = "10.1109/FG.2015.7285020",

language = "English",

isbn = "9781479960279",

volume = "4",

pages = "1--6",

booktitle = "2015 11th IEEE International Conference and Workshops on Automatic Face and Gesture Recognition (FG 2015)",

publisher = "Institute of Electrical and Electronics Engineers (IEEE)",

address = "United States",

note = "Automatic Face and Gesture Recognition (FG), 11th IEEE International Conference and Workshops ; Conference date: 04-05-2015 Through 08-05-2015",

}

Chriskos, P, Zoidi, O, Tefas, A & Pitas, I 2015, De-Identifying Facial Images Using Projections on Hyperspheres. in 2015 11th IEEE International Conference and Workshops on Automatic Face and Gesture Recognition (FG 2015): Proceedings of a meeting held 4-8 May 2015, Ljubljana, Slovenia. vol. 4, Institute of Electrical and Electronics Engineers (IEEE), pp. 1-6, Automatic Face and Gesture Recognition (FG), 11th IEEE International Conference and Workshops, Ljubljana, Slovenia, 4/05/15. https://doi.org/10.1109/FG.2015.7285020

De-Identifying Facial Images Using Projections on Hyperspheres. / Chriskos, P.; Zoidi, O.; Tefas, A.; Pitas, Ioannis.

2015 11th IEEE International Conference and Workshops on Automatic Face and Gesture Recognition (FG 2015): Proceedings of a meeting held 4-8 May 2015, Ljubljana, Slovenia. Vol. 4 Institute of Electrical and Electronics Engineers (IEEE), 2015. p. 1-6.

Research output: Chapter in Book/Report/Conference proceeding › Conference Contribution (Conference Proceeding)

TY - GEN

T1 - De-Identifying Facial Images Using Projections on Hyperspheres

AU - Chriskos, P.

AU - Zoidi, O.

AU - Tefas, A.

AU - Pitas, Ioannis

PY - 2015/8

Y1 - 2015/8

N2 - A major issue that arises from mass visual media distribution in modern video sharing, social media and cloud services, is the issue of privacy. Malicious users can use these services to track the actions of certain individuals and/or groupsthus violating their privacy. As a result the need to hinder automatic facial image identification in images and videos arises. In this paper we propose a method for de-identifying facial images. Contrary to most de-identification methods, thismethod manipulates facial images so that humans can still recognize the individual or individuals in an image or video frame, but at the same time common automatic identification algorithms fail to do so. This is achieved by projecting the facial images on a hypersphere. From the conducted experimentsit can be verified that this method is effective in reducing the classification accuracy under 10%. Furthermore, in the resulting images the subject can be identified by human viewers.

AB - A major issue that arises from mass visual media distribution in modern video sharing, social media and cloud services, is the issue of privacy. Malicious users can use these services to track the actions of certain individuals and/or groupsthus violating their privacy. As a result the need to hinder automatic facial image identification in images and videos arises. In this paper we propose a method for de-identifying facial images. Contrary to most de-identification methods, thismethod manipulates facial images so that humans can still recognize the individual or individuals in an image or video frame, but at the same time common automatic identification algorithms fail to do so. This is achieved by projecting the facial images on a hypersphere. From the conducted experimentsit can be verified that this method is effective in reducing the classification accuracy under 10%. Furthermore, in the resulting images the subject can be identified by human viewers.

U2 - 10.1109/FG.2015.7285020

DO - 10.1109/FG.2015.7285020

M3 - Conference Contribution (Conference Proceeding)

SN - 9781479960279

VL - 4

SP - 1

EP - 6

BT - 2015 11th IEEE International Conference and Workshops on Automatic Face and Gesture Recognition (FG 2015)

PB - Institute of Electrical and Electronics Engineers (IEEE)

T2 - Automatic Face and Gesture Recognition (FG), 11th IEEE International Conference and Workshops

Y2 - 4 May 2015 through 8 May 2015

ER -

Chriskos P, Zoidi O, Tefas A, Pitas I. De-Identifying Facial Images Using Projections on Hyperspheres. In 2015 11th IEEE International Conference and Workshops on Automatic Face and Gesture Recognition (FG 2015): Proceedings of a meeting held 4-8 May 2015, Ljubljana, Slovenia. Vol. 4. Institute of Electrical and Electronics Engineers (IEEE). 2015. p. 1-6 https://doi.org/10.1109/FG.2015.7285020