In years past, facial recognition software depended on good lighting for it to function properly. This made facial recognition a futile exercise when used in dark surroundings or instances when lighting wasn’t optimal. In 2018, the software landscape is changing, and the future of facial recognition on your phone will only look to become more advanced in the future.
It’s called Light Invariant Video Imaging (LIVI), and it eliminates background light, dynamic light and other issues to deliver shadow-free, clear images. The result is that images become more recognizable, making facial recognition snappy, smooth and reliable.
The LIVI system uses amplitude-modulated (AM) light separation to eliminate background and dynamic lighting. This presents the image in better contrast and without shadows. The AM light separation process works similarly to AM radio communication. In principle, it tunes out unwanted light sources like a radio would tune out radio signals.
It’s easy to see the practicality of the LIVI system, and how easily it can be applied to commonly used technologies. Smartphone cameras have long been using facial recognition systems, but most honest attempts at the software have been unsuccessful. The ability to filter backlight and other unwanted elements could be a gamechanger for the industry, as smartphone companies are looking to integrate more functional and reliable software into their devices.
The Future for Facial Recognition?
Companies like Apple have vastly improved their facial recognition systems in the past year with tools like FaceID. The future of the industry is trending toward being dominated with facial recognition—whether it’s to unlock a home screen or to purchase an app.
The worldwide smartphone market reached 1.53 billion units in 2017, and it’s safe to say the number will steadily climb in the years to come. Facial recognition is making strides to become a more secure user authentication system. System’s like Apple’s FaceID have proven to be largely invulnerable biometric hacking attempts, like using a person’s photos to spoof the identification system. Software like LIVI hopes to add to the capabilities of systems like FaceID to improve security and reliability in different lighting conditions.
The future of facial recognition is wide open, and the team behind the LIVI system hopes to be at the forefront of it. The BGU team is currently seeking additional funding for research, so it’ll likely be some time before their system reaches mainstream smart devices (if it does at all), but it’s conceivable that their software can be integrated into the devices of the future.