Researchers Hack Faces In Biometric Facial Authentication Systems

A Vietnamese researcher will demonstrate at Black Hat DC next week how he and his colleagues were able to easily spoof and bypass biometric systems that authenticate users by scanning their faces.

The researchers cracked the biometric authentication embedded in Lenovo, Asus, and Toshiba laptops by spoofing the biometric systems with everything from a photo of the authorized user to brute-force hacking using fake facial images. They successfully bypassed Lenovo's Veriface III, Asus' SmartLogon V1.0.0005, and Toshiba's Face Recognition 2.0.2.32 -- each set to its highest security level -- demonstrating vulnerabilities in the systems that let an attacker cheat them with phony photos of the legitimate user and gain access to the laptops.

These Windows XP and Vista laptops come with built-in webcams that work with the facial-recognition technology. This form of authentication is considered more convenient than fingerprint scans and more secure than traditional passwords. The software scans the user's face and stores the images and facial characteristics. Then the user can log in by scanning his or her face, which is then matched against the image data.

The researchers were able to bypass the authentication system not only by using a photo of the authorized user, but also by creating multiple phony facial images. "The mechanisms used by those three vendors haven't met the security requirements needed by an authentication system, and they cannot wholly protect their users from being tampered," the researchers wrote in their paper on the hack.

One of the researchers, Nguyen Minh Duc, manager of the application security department at the BKIS (Bach Khoa Internetwork Security Center) at Hanoi University of Technology, will demonstrate the hack at Black Hat, as well as the tool he and his colleagues developed.

"There is no way to fix this vulnerability," Duc says. "Asus, Lenovo, and Toshiba have to remove this function from all the models of their laptops ... [they] must give an advisory to users all over the world: Stop using this [biometric] function."

An attacker can edit and adjust the lighting and angle of a phony photo to ensure the system will accept it, according to the researchers. "Due to the fact that a hacker doesn't know exactly how the face learnt by the system looks like, he has to create a large number of images...let us call this method of attack 'Fake Face Bruteforce.' It is just easy to do that with a wide range of image editing programs at the moment," they wrote in their paper.

"One special point we found out when studying those algorithms is that all of them work with images that have already been digitalized and gone through image processing. Consequently, we think that this is the weakest security spot in face recognition systems, generally, and access control system of the three vendors, particularly."

Have a comment on this story? Please click "Discuss" below. If you'd like to contact Dark Reading's editors directly, send us a message