Advertisement
Advertisement
Future tech
Get more with myNEWS
A personalised news feed of stories that matter to you
Learn more
Baidu CEO Robin Li speaks to media as he arrives for the opening session of the Chinese People's Political Consultative Conference in Beijing on March 3, 2018. Photo: AFP

How artificial intelligence will change the face of security in China

Future tech

Long lines and lengthy waits for round after round of security checks are di rigueur at major events in China such as the annual meetings of its legislature.

That may change in the near future with the use of facial recognition to red-flag people for further screening while allowing others to proceed, speeding up the process of letting thousands of delegates, journalists and staff into the Great Hall of the People, according to Baidu CEO Robin Li Yanhong, whose company is testing such a system at airports. 

“The measures to verify identification, such as the tight security checks during the Two Sessions, have brought inconvenience to people’s lives,” said Li, who is a delegate to the Chinese People’s Political Consultative Conference.

“Most of the time, people’s identification is done manually, which I think will be completely unnecessary in the future,” he told mainland Chinese media ahead of the event. 

Li’s comments highlight the growing use of facial recognition in security applications in China, part of the nation’s wider push to lead the world in artificial intelligence and a desire by the central government to improve public safety through surveillance of citizens.

Currently, physical security checks begin at the subway stations, where commuters are given pat-downs and their belongings X-rayed. Paramilitary police and other security personnel man various checkpoints leading to the Hall, each time checking credentials to match face to ID.

This year’s Two Sessions also features a system dubbed “Sky Net” that sends an alarm to authorities if a wanted person is identified by facial recognition, according to a report in the state-owned Science and Technology Daily on March 2.

“This sounds like the plot of the sci-fi movie Minority Report, but it is now becoming part of people’s daily life,” said the report.

Military delegates arrive for the opening session of the National People's Congress (NPC) in Beijing on March 5, 2018. Photo: Reuters

Baidu’s Li envisions a future where AI-based technology will be advanced enough to identify “who you are” within seconds.

 “At the Palace Museum or in Tiananmen Square, whether it is 50,000 travellers or 80,000, the identification will be done by machines with a very high accuracy rate, and unlike people, machines never get tired,” he said, referring to the former palace grounds of China’s last imperial dynasty and the massive square in front of it. 

“We have been talking to the Palace Museum recently. Their biggest headache is they don’t know for sure how many of its daily visitors have left the grounds when the museum closes, worrying that some may try to stay on for a poke-around after hours,” he said.

Despite rising privacy concerns, facial recognition systems have been extensively used by Chinese authorities to spot suspected criminals and even jaywalkers. 

Some of the busiest subway stations in Beijing have already installed facial recognition cameras while railway police in central China’s Henan province have been equipped with smart glasses with built-in facial recognition technology to screen passengers. 

At the Two Sessions meetings, police officers and even police dogs were spotted wearing panoramic-view body cameras, which are linked to facial recognition software that can identify wanted suspects in real time. 

Beijing-based Nebula Science and Technology, which partnered with a research institute backed by China’s Ministry of Public Security to develop the cameras, said the devices are expected to be used by up to 50,000 police officers across the country this year.

“The cameras can spot suspects within seconds and alert police officers in almost real time,” said Nebula chief executive Shi Pengfei.

Despite the increasing adoption of advanced AI technology in security, Shi doesn’t think machines will completely replace humans when it comes to security, especially for high-profile events like the Two Sessions. 

“The efficiency of security checks can be improved dramatically with technology and the procedure will be simplified with fewer security personnel, but what if some people fake their looks or even finger prints? I think we will always need a human security presence no matter how much technology advances,” he said.

Post