Controversial Chinese "gait recognition" technology to identify people through their WALK

Controversial Chinese "gait recognition" technology to identify people through their WALK

The Chinese authorities have begun to introduce a new monitoring tool: the "gait detection software", which uses people's body shapes and their ways to identify them, even when their faces are hidden from cameras.

Gang detection has already been used by police on the streets of Beijing and Shanghai and is part of a push across China to develop artificial intelligence and data-driven surveillance.

However, the critics were concerned about how far the technology will go.

Scroll down for video

China is using a new monitoring tool: a "gait detection software" that uses people's body shape and how they go to identify them, even when their faces are hidden. Pictured is Huang Yongzhen, CEO of Watrix, the company behind the controversial technology

China is using a new monitoring tool: a "gait detection" software that uses people's body shape and style to identify them, even when their faces are hidden. Pictured is Huang Yongzhen, CEO of Watrix, the company behind the controversial technology

HOW DOES IT WORK?

Watrix's Chinese startup software extracts a person's silhouette from a video and analyzes the movement of the silhouette to create a model for the person's movement.

It is not yet able to identify people in real time.

Users must upload a video to the program instead. This takes about 10 minutes to search an hour of video.

No special cameras are required – the software can use surveillance camera footage to analyze the aisle.

According to Startup, the software has an accuracy of around 94 percent.

Huang Yongzhen, the CEO of Watrix, said his system can identify people up to 50 meters away, even with their backs turned or face covered.

This may close a gap in facial recognition that requires close-up close-ups of a person's face.

"You do not need people's cooperation so we can recognize their identity," Huang said in an interview in his Beijing office.

"The gait analysis can not be faked by simply limping, walking or stooping, as we analyze all the characteristics of an entire body."

Watrix announced last month that, according to Chinese media reports, it raised $ 14.5 million to accelerate the development and sale of its aisle recognition technology.

The Chinese police use facial recognition to identify people in masses of people and Nab-Jaywalker, and develop a national national system of surveillance camera data.

But not everyone is familiar with the use of gait recognition.

Gang detection has been used by police on the streets of Beijing and Shanghai and is part of a major effort to develop artificial intelligence and data-driven surveillance throughout China

A screen shows a computer-generated image of a Watrix employee walking in his company's Beijing offices during a demonstration of his company's motion detection software

A screen shows a computer-generated image of a Watrix employee walking in his company's offices in Beijing during a demonstration of his company's motion detection software

Security officials in China's far western province of Xinjiang, a region whose Muslim population is already under intense surveillance and control, have expressed interest in the software.

Shi Shusi, a Chinese columnist and commentator, says that it is not surprising that technology arrives faster in China than the rest of the world, as Beijing values ​​social control.

"The use of biometric recognition to maintain social stability and manage society is an unstoppable trend," he said. "It's a great deal."

In addition, the technology is not new: scientists in Japan, the United Kingdom, and the US Defense Information Systems Agency have been researching gait recognition for over a decade. They are trying to overcome different ways of overcoming skepticism that people might be recognized for walking.

Professors at Osaka University have been working with the National Police Agency in Japan since 2013 to use pilot-based gear recognition software.

But only a few have tried to commercialize the gear recognition.

Watrix employees demonstrate the gear recognition software of their company. A Chinese technology startup hopes to start selling software that recognizes people based on their body shape and walking behavior, and enables identification when faces are hidden from cameras

Watrix employees demonstrate their company's gear recognition software. A Chinese technology startup hopes to start selling software that recognizes people based on their body shape and walking behavior, and enables identification when faces are hidden from cameras

CAN YOU PUNISH THE HOW YOU GO?

Research from Southampton University has shown that criminals can be distinguished by walking.

According to their study, an individual's gait is so pronounced that it could be used for identification.

For criminals, investigators may be able to compare images of suspects being held on tape with the way they run.

Using a biometric tunnel with 123 cameras, the scientists created 3D images of 25 different people.

They believe that the airport's immigration halls could be equipped with tunnels with cameras that identify passers-by.

This means that criminals could still be identified if they masked their face or wore gloves with 95 percent accuracy.

Israel-based FST Biometrics closed earlier this year as the company struggled with technical difficulties with its products, said former board member Gabriel Tal.

"It's more complex than other biometric data, arithmetically," said Mark Nixon, a leading gait expert at the University of Southampton in the UK.

"Larger computers are required to go because you need a sequence of images instead of a single image."

Watrix's software extracts the silhouette of a person from a video and analyzes the movement of the silhouette to create a model for the person's movement.

It is not yet able to identify people in real time. Users need to upload a video to the program, which takes about 10 minutes to search an hour of video.

No special cameras are required – the software can use surveillance camera footage to analyze the aisle.

Watrix's software extracts the silhouette of a person from a video and analyzes the movement of the silhouette to create a model for the person's walking. Pictured is a demonstration

Watrix's software extracts the silhouette of a person from a video and analyzes the movement of the silhouette to create a model for the person's movement. Pictured is a demonstration

Although the software is not as good as facial recognition, Huang said the accuracy rate of 94 percent is sufficient for commercial use. He hopes it can be used alongside face recognition

Although the software is not as good as facial recognition, Huang said the accuracy rate of 94 percent is sufficient for commercial use. He hopes it can be used alongside face recognition

Although the software is not as good as facial recognition, Huang said the accuracy rate of 94 percent is sufficient for commercial use.

It provides that gait detection is used in addition to facial scanning software.

In addition to monitoring, Huang says, gait detection can also be used to detect people in need, such as elderly people who have fallen.

Nixon believes that technology can make life safer and more comfortable.

"People still do not realize that they can be recognized by their walk, while everyone knows that they can be recognized by your face," said Nixon.

"We believe you are absolutely unique."

Leave a comment

Send a Comment

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.