Is your face recognition spoof proof?

While there is debate on where, how and why to use facial recognition, many a city's are already leading by example and have started to use the technology for enforcement, banking systems (secondary authentication, business use cases). Mobile devices are now inbuilt with face recognition replacing traditional password mechanism. Systems are designed to work smoothly without users having to know their existence. As we start to become dependent on these facial recognition systems, questions remains if these systems can be fooled?

Today there are attempts to fool these systems by,

• Use of an image (through mobile or ID) 
• Use of 3D mask
• Use of a video in 3D space
• Confusing system by rapid movements
• Changing expressions, eye movements etc.

The major advantage of face recognition system is its accuracy and smoothness. User dose not even have know that they are being authenticated  by any machine. But due to above mentioned challenges even top players in the face recognition field can fail. It is evident that we need carefully drafted deep learning algorithms to detect spoofing attempts and the solutions that incorporate number of camera's to create a layer of security.