FR-CAPTCHA: CAPTCHA Based on Recognizing Human Faces
A Completely Automated Public Turing test to tell Computers and Humans Apart (CAPTCHA) is designed to distinguish humans from machines. Most existing tests require solving distorted text embedded in a background image. Many existing CAPTCHAs are either too difficult for humans due to excessive distortions or trivial for automated algorithms. These CAPTCHAs also suffer from inherent language as well as script dependency and are not equally accessible for people belonging to varying demographics. However, there can be other Turing tests that encompass all human races and demographics. One such test is matching two faces and establish if they belong to same individual or not. Utilizing face recognition capabilities as the Turing test, we propose FR-CAPTCHA based on finding matching pairs of human faces in an image. We observe that, compared to existing implementations, FR-CAPTCHA achieves higher human accuracy and is robust towards automated algorithms.
How to solve a FR-CAPTCHA: Each FR-CAPTCHA image has multiple face images embedded in it. The user has to mark the location of two matching faces (i.e. face images belonging to the same individual) to solve the CAPTCHA. Selecting a non-face image or non-matching face image will result in an incorrect response.
For a complete sample set of CAPTCHAs with their ground truth data, please contact: gauravgs@iiitd.ac.in
