A New Captcha Protocol For Avoiding Machine Learning Attacks
Abstract— Many Different techniques are used to against AI problems. We introduced a new concept for providing more security in online through the CARP - Captcha as a graphical password). We have proposed to combine the Captcha and graphical password to provide more security. CARP is preventing the various bots such as e-mail attacks, online guessing attacks, relay attacks, shoulder surfing attack, dual view technology using Hash techniques. A Novel Approach is provided by CARP to detect the well-known image hotspot problem in most graphical password systems such as PassPoints. Recall based scheme is used by the Image Recognizer to generate the CAPTCHA. CARP is motivating new inventions such as machine learning attacks. If the password is in the search set, a CARP password can detect only probabilistically by routine online guessing attacks. Through TLS (Transport Layer Security) the CARP schemes are used with additional protection such as a secure channel between the clients and the authentication server.
Index Terms — CARP, E-mail attacks, Online guessing attacks, Relay attacks, Shoulder surfing attack.
Click Here
International Journal for Trends in Technology & Engineering © 2015 IJTET JOURNAL