[youtube https://www.youtube.com/watch?v=V_F_z1p0qSY]
Machines are getting increasingly impressive at recognizing things — most of the time. Occasionally, some reminder comes along which showcases how something that’s instantly recognizable to humans can, with a few tweaks, be rendered totally baffling to a computer. Capitalizing on this, an Israeli company has developed some smart software that promises to help crack down on facial recognition technology being used against people in a way they may not be happy with.
Founded by former members of the Israeli Defense Forces, D-ID‘s tech works by replacing faces in video or still images with near-identical looking photorealistic portraits. While these faces look virtually indistinguishable from the real person to fellow humans, they are non-identifiable to facial recognition technologies.
“The way our technology works is that we remove the facial image without processing or profiling the individual, replacing it with an A.I.-generated image of a nonexistent person that actually looks like the same person to the human eye,” Gil Perry, co-founder and CEO of D-ID, told Digital Trends. “By using an A.I.-generated image, this means that the face cannot be identified through techniques like decryption or reverse engineering, so it outsmarts even the most sophisticated facial recognition engines, while respecting privacy laws and regulations. Our technology is also far superior to traditional methods of image or video protection — whether that’s encryption, pixelation, or blurry — which have been proven to not fully protect people’s identities.”
This isn’t just something that would be of benefit only to privacy-conscious individuals. Companies can use it, too, ensuring that they get to offer the benefits of facial recognition without the underlying creepiness of the tech. For example, a growing number of stores are interested in the idea of using smart cameras to identify customers for things like checkout-free automated payments. But capturing a person’s face has privacy implications. D-ID’s technology could therefore be used to anonymize face data in real time to avoid these challenges.
The same is true with technologies like in-car driver monitoring in cars. Companies could use this anonymized footage to extract data analysis and statistics — but without the fear that the data could potentially be stolen and repurposed to, for instance, track down individuals.
“While we were in the [Israeli Defense Forces] we could not have any photos taken of us — and certainly not posted or shared on social media networks,” Perry continued. “We recognized the growing global use of facial recognition technology and the challenges it posed to user privacy and anonymity. This led to us developing the first facial image de-identification solution, which aims to protect privacy: what is now D-ID today. Our system is currently being used in leading Fortune 500 companies and organizations worldwide.”