face recognition

Facial recognition systems as a security protocol have grown in presence in recent years, mainly thanks to their implementation on some mobiles.

Like all technology, this system has certain base vulnerabilities. Depending on the robustness of each system, these could be violated with photographs or using the face of the lock holder in unforeseen situations, such as while sleeping. To address these weaknesses, a new alternative was presented.

Facial unlocking with gestures, as a two-step reinforced option

This proposal was presented by Brigham Young University Professor of Electrical and Computer Engineering, D.J. Lee, who says there is a better and more reliable one to use a face as an access mechanism for restricted control.

El mecanismo se llama C2FIV, sigla de Concurrent Two-Factor Identity Verification (‎verificación simultánea de identidad de dos factores‎, en español). To validate an unlock order, the system recognizes the facial identity of the person in front of the camera and a specific movement or gesture.

To set up this unlocking system, the user must stand in front of the camera and record a short one- or two-second video of a specific movement with the face or lips, reading a secret phrase. The video is then processed on the device, where facial features are analyzed, plus the trajectory of face movement, and stored for further identity verification.

In a preliminary study, Professor Lee and his Ph.D. student Zheng Sun recorded 8,000 video clips containing facial movements of 50 subjects. These movements included flickering, jaw movements, smiling or raising eyebrows, and many other random facial gestures to train the neural network. They then produced a set of positive and negative facial motion pairs and entered a higher score for the positive pair (paired pair between the request made and the previously configured records).

C2FIV relies on an integrated neural network framework to simultaneously learn facial features and actions. The framework models dynamic and continuous data, such as facial movement, where all recorded frames should be considered, unlike still photos that can describe people.

face recognition

With this integrated neural network framework, the user's facial actions and features can be embedded and stored on a server or device. Once enabled, when the system receives unlock requests, the computer will compare the newly generated embeddings with those stored in its database to validate the requests. The user's identity verification process is governed by a predefined threshold for assessing similarities between new and previously stored additions.

Lee has already patented this technology. According to his own statements, the idea behind this project is not to compete with Apple, nor to seek for this system to be implemented on smartphones. According to the plans of its creator, C2FIV's vocation is broader, including access to restricted areas in a job establishment, to log in to online banking, for the use of ATMs and even, to dispense with a car key.

Post a Comment