Visible to the public A Text-Based CAPTCHA Cracking System with Generative Adversarial Networks

TitleA Text-Based CAPTCHA Cracking System with Generative Adversarial Networks
Publication TypeConference Paper
Year of Publication2018
AuthorsLiu, F., Li, Z., Li, X., Lv, T.
Conference Name2018 IEEE International Symposium on Multimedia (ISM)
Date Publisheddec
KeywordsCAPTCHA, captchas, cDCGAN, CNN, Completely Automated Public Turing Test to Tell Computers and Humans Apart, composability, computer security, conditional deep convolutional generative adversarial networks, convolutional neural nets, convolutional neural networks, Fans, generative adversarial networks, Human Behavior, multimedia computing, multimedia security mechanism, Multimedia systems, Pearson correlation coefficients, pubcrawl, security of data, small sample problem, text analysis, text-based CAPTCHA, text-based CAPTCHA cracking system, Training
AbstractAs a multimedia security mechanism, CAPTCHAs are completely automated public turing test to tell computers and humans apart. Although cracking CAPTCHA has been explored for many years, it is still a challenging problem for real practice. In this demo, we present a text based CAPTCHA cracking system by using convolutional neural networks(CNN). To solve small sample problem, we propose to combine conditional deep convolutional generative adversarial networks(cDCGAN) and CNN, which makes a tremendous progress in accuracy. In addition, we also select multiple models with low pearson correlation coefficients for majority voting ensemble, which further improves the accuracy. The experimental results show that the system has great advantages and provides a new mean for cracking CAPTCHAs.
Citation Keyliu_text-based_2018