BiometricJammer: Preventing fingerprint replication from camera images | PrivacyVisor: Protection against facial recognition from camera images | Gait anomymization
BiometricJammer: Preventing fingerprint replication from camera images
- T. Ogane and I. Echizen, "BiometricJammer: Preventing surreptitious fingerprint photography without inconveniencing users," Proc of the International Joint Conference on Biometrics 2017 (IJCB2017),pp.253-260,(October 2017)
- I. Echizen and T. Ogane, “BiometricJammer: Method to prevent acquisition of biometric information by surreptitious photography on fingerprints,” IEICE Trans. on Information & Systems, Vol.E101-D, No.1,pp.2-12, (January 2018) [Invited paper], Link
- T. Ogane and I. Echizen, “BiometricJammer: Use of pseudo fingerprint to prevent fingerprint extraction from camera images without inconveniencing users,” Proc. of the 2018 IEEE International Conference on Systems, Man, and Cybernetics (SMC2018), 7 pages, (October 2018)
With the increase in the number of pixels in digital cameras, fingerprint information that could previously only be read by contact fingerprint sensors can now be recovered from images taken by digital cameras, and the danger of it being misused for unauthorized logins and identity theft has been pointed out.
In October 2016, we showed that it is possible to extract fingerprint information necessary for fingerprint authentication from finger images taken from a distance of 3 m with a commonly available digital SLR camera, and then proposed a fingerprint stealth prevention method, BiometricJammer, to prevent such extraction. This is a method in which a pattern (jamming pattern) designed to interfere with the detection of fingerprint feature points is transferred to the fingertip using a stencil sheet, preventing the extraction of fingerprint information from the image of the fingertip with the jamming pattern attached. On the other hand, even with the jamming pattern attached, the contact fingerprint sensor will still recognize the fingerprint normally.
Reference
PrivacyVisor: Protection against facial recognition from camera images
- BBC News(UK), Privacy visor blocks facial recognition software January 22, 2013
- NBC News, LED-powered 'privacy visor' thwarts facial recognition June 20, 2013
- TIME, Leery of Facial Recognition? These Glasses Might Help June 20, 2013
- T. Yamada, S. Gohshi, and I. Echizen, "Use of invisible noise signals to prevent privacy invasion through face recognition from camera images," Proc. of the ACM Multimedia 2012 (ACM MM 2012), pp.1315-1316, (October 2012)
- T. Yamada, S. Gohshi, and I. Echizen, “Privacy Visor: Method for Preventing Face Image Detection by Using Differences in Human and Device Sensitivity,” Proc. of the 14th Joint IFIP TC6 and TC11 Conference on Communications and Multimedia Security (CMS 2013), 10 pages, (September 2013)
- T. Yamada, S. Gohshi, and I. Echizen, “Privacy Visor: Method based on Light Absorbing and Reflecting Properties for Preventing Face Image Detection,” Proc. of the 2013 IEEE International Conference on Systems, Man, and Cybernetics (IEEE SMC 2013), 6 pages, (October 2013)
The problem of easy disclosure of privacy information has become apparent with the spread of mobile terminals with built-in sensors such as cameras and GPS, and the development of facial recognition technology. If a photo taken without permission or unintentionally appears in a picture and is disclosed on a social networking service (SNS), the information about when, where, and with whom the person concerned was exposed through the face recognition function. There is a risk that one's name, place of work, hobbies, etc. may become known through the facial recognition function just by walking on the street, and there is a need for essential countermeasures to prevent privacy violations caused by camera capture. Against this background, we have developed the world's first technology to prevent privacy violations caused by voyeurism and camera capture (PrivacyVisor). By having the subject wear PrivacyVisor, it is possible to fail to detect the subject's face without adding any new functions to the existing camera.
Reference
Gait anomymization
- Tieu, Ngoc-Dung T., et al. "An approach for gait anonymization using deep learning." 2017 IEEE Workshop on Information Forensics and Security (WIFS). IEEE, 2017.
- Tieu, Ngoc-Dung T., et al. "Spatio-temporal generative adversarial network for gait anonymization." Journal of Information Security and Applications 46 (2019): 307-319.
- N. Tieu, J. Yamagishi, and I. Echizen "Color Transfer to Anonymized Gait Images While Maintaining Anonymization," Proc. of the APSIPA Annual Summit and Conference 2020 (APSIPA ACS 2020), pp. 1406-1413, December 2020
There is a threat that a person's privacy may be violated if the same person in another video is identified through gait recognition based on the gait of the person in the video. To solve this problem, we proposed a gait anonymization method that makes gait recognition difficult without changing any attributes other than gait and without reducing visual degradation of the person's appearance and movements. Specifically, we studied a basic anonymization method using silhouette image sequences as input and output, an anonymization method using color images of a person walking as input and output, and an anonymization method for incomplete silhouette image sequences.