Title: A novel CAPTCHA scheme based on facial expression reconstruction

Authors: Mohammad Moradi; Mohammad Reza Keyvanpour

Addresses: Data Mining Laboratory, Department of Computer Engineering, Alzahra University, Tehran, Iran ' Department of Computer Engineering, Alzahra University, Tehran, Iran

Abstract: Despite some technical and conceptual criticism on the CAPTCHA and its claimed effectiveness, such human interaction proofs (HIPs) are still in the centre of attention. With the aim of designing a novel, efficient and user-friendly CAPTCHA, the facial expression reconstruction CAPTCHA, FERCHA, is proposed. The distinguishing challenge in this schema is reconstruction of a facial expression related to the perceived emotion in a game-like manner. In one side, the weakness of current machines in performing cognitive tasks has been leveraged to introduce a strong CAPTCHA. Also, as the byproduct, the FERCHA schema can be used as a crowdsourcing platform in order to collect human-generated semantic annotation and interpretation of text, emojis and images. The experimental results (including theoretical analysis and user acceptance evaluation) proved the efficacy and robustness of the presented idea. Moreover, it is found that providing humans with user-friendly experience could improve the success rate and satisfaction level.

Keywords: human interaction proof; HIP; CAPTCHA; web security; facial expression recognition; emotion detection; human computer interaction; sentiment analysis; semantic interpretation; gamification.

DOI: 10.1504/IJEB.2020.111061

International Journal of Electronic Business, 2020 Vol.15 No.4, pp.368 - 388

Received: 08 Apr 2019
Accepted: 14 Aug 2019

Published online: 21 Oct 2020 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article