Snapchat is a silly thing.
13- to 18-year-olds use it to send goofy photos and videos to each other. Brands use it to sell their products to those same 13- to 18-year-olds. Regardless of whether you're a brand, part of the audience, or something else entirely, there’s no denying that some pretty powerful technology is at work inside that little app, and it's been showing up elsewhere, too.
The basics of facial recognition
Facial recognition software pairs highly complex computer vision technology with fine-tuned algorithms to map a face using a camera. For starters, it can simply scan an image to recognize the basic shape and outline of the face. The software can also scan an image for light and dark patterns that are programmed to be recognized as a face. It can tell things like where your nose and forehead are and how far apart your eyes are.
This is where it begins to get very dynamic.
Not from scratch
Programs and apps that use facial recognition software aren’t necessarily scanning a new face from scratch every single time. They often begin with a composite of hundreds (or maybe thousands) of pre-mapped faces, allowing the image scanner to begin with a face template, making specific adjustments more quickly and easily.
This is how apps like Snapchat (which has invested heavily in the technology in the last year) can recognize specific facial features and provide interactive lenses. This video from Vox explains it perfectly.
Like most advances in the semiconductor industry, its most basic applications pale in comparison to their potential. Here are three applications with the biggest impact:
Banks and credit card companies
As mobile rockets into prominence, financial institutions grapple with maintaining robust security while users are on the go. Integrating facial recognition software into their mobile apps is a promising way to phase out hackable passwords and integrate security features that are extremely difficult to forge.
Whales have faces, too.
The North Atlantic right whale is “…one of the most endangered of all whale species,” according to The Atlantic. Researchers are working hard to protect the few that are left and it seems as though quicker facial recognition is helping. An aerial fly team has begun using facial recognition technology to better track and monitor individual whales in targeted pods as they travel and feed.
Researchers and doctors at the University of Oxford have developed their own kind of facial recognition software, with their sights set on better understanding rare genetic diseases. “Around 30-40 percent of genetic disorders, including Down’s syndrome and Marfan syndrome," according to the Medical Research Council, “involve changes to the face or skull.” Their use of facial recognition is helping them to pinpoint “new disorders and the DNA variations that cause them.”
This technology is over 10 years old, and what’s new is that smaller mobile devices are now equipped with enough processing power to run this technology in real time. As long as the semiconductor industry can keep driving that trend, the capabilities will continue to amaze us.