The best way to identify someone is by the way they walk, not their face

The Rundown: 

Gait recognition technology uses surveillance cameras to identify individuals by the way they walk. And it’s unbelievably accurate. This is why the Chinese government is implementing the technology to identify and watch their citizens. But there are some other interesting use cases, such as analyzing running form. By 2026, digital running coaches will use gait recognition technology to analyze and improve the form of casual runners.

Did you know that your stride is as unique as your fingerprint? Your walking style is an identifying characteristic just as much as your DNA. Sounds absurd, right?

I heard a story five years ago about how FBI operatives were trained to recognize a target based solely on the way they walked. Basically, the idea was that everyone had a gait (stride) that was unique to them. And if you learned someone’s gait, then you could easily identify them amongst a crowd, even if they were wearing a disguise.

I’ll be honest, I was skeptical. It didn’t really make sense. What if you purposely limped or slowed down? Wouldn’t this throw it off?

Well, as it turns out, gait recognition is a real thing. And the technology is well on its way.

Already a future thinker?
Then become a friend.

Gait Recognition

We’ve all heard of facial recognition technology being used to identify people via surveillance cameras. Well, Watrix is a company in China that is taking a different approach through technology known as gait recognition software.

Gait recognition software uses people’s body shapes and how they walk to identify them, even when their faces are hidden from cameras. Watrix, extracts a person’s silhouette from video and analyses the silhouette’s movement to create a model of the way the person walks.

It doesn’t require special cameras — the software can use footage from surveillance cameras to analyse gait… from up to 50 metres away, even with their back turned or face covered.

ABC News

Currently, gait recognition is a little on the slow side of things. It cannot be done real-time because the system must analyze videos instead of still images. Nonetheless, the Chinese government has already shown interest in this technology and it will likely play a role in their social credit system they’re deploying in 2020.

If you’re not familiar, the Chinese Social Credit System is a mass surveillance plan, whereby, they’re standardizing the assessment of citizens’ and businesses’ economic and social reputation, or credit. It’s a way of preventing people from talking poorly of the government, breaking laws, etc.

Here in America, the media has described it as the start of a digital dictatorship. Without getting too far into the details, I think we need to consider the cultural differences before we jump to these conclusions. Very simply, the Chinese see privacy differently than Americans.

What I find particularly fascinating is whether or not the gait recognition software can be fooled.

Can it be Fooled?

I’ve talked in great detail about digital disguises in the past and how we can in fact confuse facial recognition algorithms. However, the researchers behind Watrix are confident that their technology is foolproof:

Gait analysis can’t be fooled by simply limping, walking with splayed feet or hunching over, because we’re analysing all the features of an entire body.

Huang Yongzhen, CEO Watrix

In an interview, Former Chief of Disguise for the CIA, Jonna Mendez, mentioned a similar argument. She said that mentally deciding to walk differently, with a limp for example, is not enough to disguise your gait. She suggested placing a piece of gravel in her shoe as a way to truly change her gait.

I’m really torn on gait recognition technology. The CIA knows how to disguise operatives and I’d like to believe that their method would disguise our gait. However, there’s no convincing evidence that this is true. Not to mention, this parameter has probably already been accounted for within the Watrix technology. After all, they have a 94% successful identification rate.

I think it’s far too early to tell whether or not gait recognition software can be fooled. There’s not much research publicly documented about it. Today, the battles are mostly being fought over facial recognition technology.

I do, however, want to point out that gait recognition may bring some good to our society outside of surveillance.

The Health Gait

Dina Katabi and her research team at MIT have created a stationary device in the home that will disrupt all senior medical alert systems. Using radio waves with 1/1000th the power of your wifi signal, it can peer through your walls and track your movements – all without a single camera.

The device is called WiGait and it operates by bouncing and receiving a signal off of your body  to measure gait velocity – what doctors sometimes call the “sixth vital sign”.

The gait velocity, as you may have guessed, is a determined by how fast someone walks, coupled with their stride length. In other words, it’s a non-visual description of movement. For that reason, gait velocity has always been a great indicator of elderly mobility.

While most medical alert systems require the senior to press a button to alert authorities, the WiGait is a system that would understand a person’s normal gait velocity, and alert help when there were abstractions to their normal gait (i.e. they fall).

There’s even a possibility that gait sensors would be used in residential spaces through the means of smart flooring and carpet – helping to correct our poor gait behaviors. One, they would notice when we’re getting too lazy and suggest we get active. Two, the flooring could detect abnormalities in one’s gait, which are the warning signs of many problems involving the spine, heart, and brain.

In speaking of gait abnormalities, perhaps the most effective use of gait recognition is in running.

Runner’s Gait

Gait analysis has been around competitive and casual running for years. Unfortunately, it’s always been a bit exclusive because it takes a professional, trained eye to analyze a stride and give meaningful insight.

However, emerging technology, such as KinetiGait, use computer vision to capture and analyze a runner’s motion. In other words, what used to take a trained professional can now be done with this application. Not to mention, not a single data point goes unnoticed and not a single insight goes overlooked.

At the moment, physical trainers and biomechanists are the only ones that can make sense of the data derived from a 3D gait recognition system like KinetiGait. This means that the competitive world of track & field will likely be the first to realize the potential of this technology. This is no surprise, considering the importance of Sports Science at the Olympic level. But, the possibilities are much greater than professionals.

In fact, the KinetiGait software is compatible with the Microsoft Kinect 2.0 camera (this is what they pair with Xbox consoles), which indicates that they have their sights set on the casual running market.

In just a few years, gait recognition systems will sync with voice assistants to create digital running coaches that can deliver guidance to everyday people via an in-home 3D gait analysis. The key here is making it simple for the average jogger to interact with a digital running coach – and making the insights easy to apply to one’s life.

These digital running coaches exist already, such as Lifebeam VI. And I’ve even used them to improve my running abilities. However, they lack the computer vision component where you’re gait is being analyzed with a fine-tooth comb.

I’ve got a cousin who ran her first Chicago marathon a few years ago. The pain she experienced leading up to the marathon she assumed was just from the increased training. Turns out her form was off and the increased training exacerbated the issue. She’s now got semi-permanent damage to her hips, knees, and shins.

Ryan, I/H Member

Whether you’re a jogger or not, you probably have heard a story of someone hurting themselves from running over years of their life. This can usually be attributed to poor form – down to the way your arms swing and your feet hit the ground. With nearly 60 million casual runners and joggers in just the U.S., the market for improving people’s running form is massive.

By 2026, digital running coaches will use gait recognition technology to improve the form of casual runners.

Join Discussion