New York police use a lot of Chinese face technology? The media should wake up!

Home > Sci-Tech

New York police use a lot of Chinese face technology? The media should wake up!

2019-01-22 17:02:38 179 ℃

South China Morning Post website screenshot: New York police use Chinese face recognition technology, and now the report has been deleted

Editor: Jinjin

New York police use Chinese face recognition technology recently published in the English edition of South China Morning Post. New York police later denied the claim and pointed out many mistakes in the report. Face recognition in China is the world's leading technology, but for any developing technology, the most unnecessary thing is the media "slaughter". Chinese media should be more rigorous and accurate to create a healthy public opinion orientation and environment for the development of AI industry in China.

New York City Police Department uses Chinese Face Recognition Technology extensively: How strong is China in this respect? Recently, a report in the English edition of the South China Morning Post triggered a lot of media follow-up, one of which used the above sentence as the headline directly.

and even how good is Chinese face recognition technology? New York police use it extensively, identifying 10 million people in one second.

No matter how powerful the Chinese face recognition technology is, is the New York Police really using the Chinese face recognition technology?

South China Morning Post: The New York Police are using Haikangwei products

extensively. Last Friday, the South China Morning Post published a report saying that "the New York Police Department is using cameras and facial recognition software developed in China". The article also gives a concrete example. River Park Towers, a 44-storey building in the Bronx District of New York with high crime rate and government subsidies for low-income residents, has installed a video surveillance system of Haicongwei TV in China. A suspect was scanned by the system after entering the building and immediately triggered an alarm, which enabled the New York police to successfully arrest him.

reported that "although it is not clear at present the specific scope and extent of the use of Chinese cameras and software in New York City's security network... uuuuuuuuu But the New York Police Department is using Haiconway's products on a large scale. The report also said that face recognition technology developed in the United States is prone to errors in identifying darker-skinned Afro-Americans and Latin Americans, and that Chinese face recognition technology performs well in this respect, which is also a major reason why the New York Police have adopted Chinese face recognition technology extensively. After the publication of the report,

, there was a strong response in the American media. Eventually, the NYPD had to officially respond to VOA via e-mail: "NYPD does not use Haiconway technology, nor does it use any face recognition technology for the real-time video system we access."

So is it the South China Morning Post reporting errors or the New York Police hiding facts?

Based on the information of 2014, confusing IP camera with face recognition technology

after a close look at the Nanzao report, we found that the example of Haikongwei technology used by the New York City Police Department is a report (Cutting-edge Surveillance), which was published about five years ago by the US network security website Security Today. ) As the world's largest supplier of identification monitoring equipment,

, "It's perfectly normal for New York City Police to use Haikongwei's products in 2014," Zak Doffman, chief executive officer of digital Barriers, a foreign AIoT video surveillance equipment manufacturer, wrote in Forbes.

"But the situation is obviously different now." Doffman wrote. John Honovich, founder of IPVM, an American video surveillance research and testing company, wrote on his website (IPVM.com) that River Park Towers installed an IP Camera, which was installed in 2012 and belonged to the New York City Public Housing Economic Camera Project, not the Facial Recognition System used by the New York City Police Department. Even now, 99.999% of IP cameras sold on the market are not "AI products", but simply cameras with simple analysis, and their analysis is often inaccurate, let alone products in 2012, Honovich pointed out.

The media should report AI development professionally and objectively

Now, the original report on Nanzao's website has been deleted.

As early as three years ago, Professor Zhou Zhihua of Nanjing University pointed out that AI in China is technically normal and progresses steadily. In the media, it's overheated and chaotic.

In 2016, Professor Zhou Zhihua was elected New Chi-yuan China Artificial Intelligence Person of the Year with 48% of the votes. At that time, Professor Zhou Zhihua said in an interview with Xin Zhiyuan that he hoped that the media could describe the development of AI more professionally and objectively to the society.

"Some media irresponsibly exaggerate, resulting in the"threat theory of artificial intelligence"and other boring statements have a negative impact on the public."

"(Current) Short Board is how to improve the public and even government departments'objective understanding of the development status and prospects of AI technology." Three years later, the demand for accurate and objective reports of AI and even high-tech industries in China has not decreased, rather more. It is a good thing to report the development of AI technology in China. However, in 2019, when AI technology in China is in a critical period of development, the most unnecessary thing is the media "slaughter".

[Join the community]

< p > New Intelligence Unit AI Technology + Industrial Community Recruitment, welcome the students interested in AI technology + industrial landing, _2 join the group; after the audit, we will invite them to join the community, after joining the community, we must revise the group notes (name-company-position; professional group audits are more rigorous, please understand).