AI That Can Identify People by the Way They Walk Can Be Police States' New Toy

Artificial intelligence has been advancing at such a fast pace these days that no one can escape its influence. There are now AIs that can recognize faces and fingerprints from afar, and now, there is an AI that can identify someone just by the way they walk. Police states are about to get a new toy that they can use to keep their citizenry in check.

Called the SfootBD, this new AI isn’t actually the first to employ the technology to analyze people’s gaits and then tie them to a particular person. However, it’s the fact that its accuracy surpasses previous examples by up to 371 percent that makes it stand out. The details are published in the IEEE Transactions on Pattern Analysis and Machine Intelligence journal, where the researchers detail exactly what the technology involves.

“Human footsteps can provide a unique behavioral pattern for robust biometric systems. We propose spatiotemporal footstep representations from floor-only sensor data in advanced computational models for automatic biometric verification. Our models deliver an artificial intelligence capable of effectively differentiating the fine-grained variability of footsteps between legitimate users (clients) and impostor users of the biometric system,” the paper reads.

“We report state-of-the-art footstep recognition rates with an optimal equal false acceptance and false rejection rate of 0.7% (equal error rate), an improvement ratio of 371% from previous state-of-the-art.”

In a statement, lead author of the study, Omar Costilla Reyes from the University of Manchester, explains why the human gait can work as a means of identifying people, Gizmodo reports. Apparently, there are several ways to differentiate just how people walk.

“Each human has approximately 24 different factors and movements when walking, resulting in every individual person having a unique, singular walking pattern,” Reyes said.

Apparently, Reyes and his team had to come up with creative ways to actually develop an AI that could recognize gaits with such accuracy. It seems all of their efforts was worth it.