Avatar CAPTCHA: Telling computers and humans apart via face classification

Darryl D'Souza, University of Louisville
Phani C. Polina, University of Louisville
Roman V. Yampolskiy, University of Louisville


This paper introduces Avatar CAPTCHA, an image based approach to distinguish human users from computer programs (bots). The proposed CAPTCHA asks users to identify avatar faces from a set of 12 grayscale images comprised of a mix of human and avatar faces. Experimental results indicate that it can be solved 62% of the time by human users with an average success time of 24 seconds and a positive user rating of 90%. It is designed to be secure against computer programs (bots). Using brute force attack the success rate for a bot to solve it is 1/4096. © 2012 IEEE.