Sony T
Some of the researchers have created up with a new Artificial Intelligence tool that can read the text captcha schemes which is used to defend the majority of the world most popular websites from the cyber attacks and malicious malware.
The algorithm which is based on the deep learning methods is the most effective solver captcha security and authentication system to date and could also spell the end for one of the most widely used website security systems.
Text-based captchas use a jumble of numbers and letters, along with some of the other security features such as the occluding lines, to distinguish between the malicious automated computer and humans programmes. It relies on people finding it much more easier to decipher the characters than machines.
The tool has been developed by the Lancaster University in the UK, Northwest University in the US and Perking University in China, which delivers out the higher accuracy than the previous captcha attack system. It is also able to successfully crack version of the captcha where the previous attack system has failed.
The solver is also highly efficient and effective. It can solve the captcha just within the 0.05 seconds by using a PC. The method involves teaching a captcha generator programme to produce up with a large number of training captchas that are different between the real captchas.
“It allows an adversary to launch an attack on services, such as Denial of Service attacks or spending spam or fishing messages, to steal personal data or even forge user identities,” said Guixin Ye, the lead student author of the study.
“Given the high success rate of our approach for most of the text captcha schemes, websites should be abandoning captchas,” Ye said.
Nice Info
I totally agree with this content.
It is really helpful to people solving some difficult captcha on web browsing.
Thank you for sharing your great knowledge with us.