Compared to our other CAPTCHAs, FR-CAPTCHA incorporates face recognition (rather than face detection) to add an additional element of challenge for would-be attackers. FR-CAPTCHA combines photos of several human faces along with other images including cartoons and outdoors photographs.
Each CAPTCHA includes two pairs of photographs where the same person appears twice. To solve the CAPTCHA, the user must locate one of the pairs and click on both of the photos of the same person without any mis-clicks.
Results
In testing, humans achieved 94% accuracy in solving FR-CAPTCHA images.
Demos
Publications
Journal Articles
- G. Goswami, B. M. Powell, M. Vatsa, R. Singh, and A. Noore, “FR-CAPTCHA: CAPTCHA Based on Recognizing Human Faces,” PLoS ONE, vol. 9, no. 4, p. e91708, Apr. 2014. PDF Publisher's Website
Conference Papers
- G. Goswami, R. Singh, M. Vatsa, B. M. Powell, and A. Noore, “Face Recognition CAPTCHA,” in Proc. IEEE 5th Int. Conf. Biometrics: Theory, Applications and Systems, Washington, DC, 2012. PDF Publisher's Website
What others are saying about FR-CAPTCHA
- B. Richmond, “Bring On This Facial Recognition CAPTCHA Already,” Motherboard, 16-Apr-2014. [Online]. Available: http://motherboard.vice.com/read/bring-on-this-facial-recognition-captcha-already.