On December 19th, Se-dol Lee defeated Handol, a South Korean Artificial Intelligence (AI) program, on the first day of his retirement Go match sponsored by Bodyfriend's patented technology "Brain Massage" in South Korean. Lee is a South Korean professional Go player of 9 dan rank and former Go champion. This match brought people back to his match with AlphaGo, an AI program developed by Google. In the first match, Lee was defeated by 1:4. In 2017, Jie Ke, the most talented and youngest Chinese Go player who had 5 Champion titles, was also defeated. The battle between human Go masters and AI gained much attention and sparked a heated debate on AI technology. [1-3]
Is AlphaGo a real AI? Dr. Zeng Yi from Institute of Automation，Chinese Academy of Sciences, believes that currently AI is still databased AI and is fundamentally different from biological intelligence because we cannot identify the decision-making mechanism of human brain. Despite this, various technology giants and startups have begun to deploy AI in different applications, with remarkable achievements applied into every aspect of daily life. That is, AI is infiltrating everywhere. [4]
How does AI work? One key technology is “Deep Learning”, a concept stemming from artificial neural networks. In brief, it is to build a neural network that simulates the human brain for analysis and learning, so that enables machines to analyze, learn, and recognize data such as texts, images, and sounds. [5]
Autonomous driving is an important application of AI’s “Deep Learning” in the revolution of automobiles. In recent years, following Google, Tesla, Uber and other high-tech incorporations, various domestic and aboard auto companies have stepped into vigorous development of self-driving technology. It is to make a car act like a human. Through intelligent software and a variety of sensing devices, the car collects the information of road conditions, including locations and obstacles in front and around, and then analyzes and makes decisions such as acceleration, deceleration, braking or changing steering wheel angle. [6-7]
Autonomous driving is designed and developed originally to achieve safe and efficient driving. It will help enhance highway safety and ease traffic congestion. However, there’s still a need for improvement on reliability and safety of autonomous driving systems existing now. According to Beijing News, Tesla Autopilot and Tesla Model 3 self-driving systems have been involved in multiple traffic accidents since 2016. Tesla vehicle manuals show that under certain circumstances, it is difficult to identify stationary vehicles, which is undoubtedly a high risk. The full popularizing of autonomous driving and even realization of unmanned driving call for improvement both on AI technology and corresponding policies and regulations. From this aspect, autonomous driving still has a long way to go. [6-8]
Facial recognition is another major practical achievement of “Deep Learning” that has been applied in many fields. It is a kind of biometric authentication that identify people by their facial features. Compared with fingerprint recognition and iris recognition, it has characteristics of unperceived recognition, highly accuracy and fast recognition. Hikvision, a video-centered Internet of Things (IoT) solutions and data operation services provider, has developed a powerful near-field facial recognition access system solution, which can achieve fast and accurate face identification. According to reports, this solution can be used to solve security problems for residential communities, the financial sector, judicial institutions, the military, public security system, border inspection, government, aerospace, power industry, factories, education, medical system and many other enterprises and institutions. [9-10]
In addition, facial recognition technology has also fostered a brand-new payment approach-face scan payment, which brings more convenience to payment service. Recently, at the World Internet Conference (WIC) in China, China UnionPay and more than 60 other institutions such as the Industrial and Commercial Bank of China (ICBC)，jointly released “Face Scan Payment” service. Before them, non-bank payment institutions such as Alipay and Tenpay had already launched face scan payment services and products. [11]
Although facial recognition technology has a considerable prospect, there are risks in its actual applications, information leakage for example. Therefore, corresponding application specifications and technical standards should be introduced to ensure the safe application. And individual users need to treat and use facial recognition products with caution, make sure that personal information and other privacy are under protection. [11]
reference:
[1] Shame before snow! Li Shishi of South Korea wins AI in retirement, but this AI is not as good as alphago
December 19, 2019 prospective economist
http://finance.sina.com.cn/stock/relnews/us/2019-12-19/doc-iihnzhfz6915704.shtml
[2] Li Shishi 1:4 regrets alphago! A full review of the man-machine go battle of Netease Technology (Cui Yuxian) October 31, 2019
https://www.ximalaya.com/renwen/7822327/224279706
[3] Ke Jie cries when talking about alphago for the first time: despair shaking all over basketball funny story says May 21, 2018
https://cloud.tencent.com/developer/news/217220
[4] Global pattern of artificial intelligence: won the alphago of kerjie, is it still not artificial intelligence? 36 krypton
2019-10-15
http://baijiahao.baidu.com/s?id=1647441623639480222&wfr=spider&for=pc
[5] In depth learning Baidu Encyclopedia
https://baike.baidu.com/item/%E6%B7%B1%E5%BA%A6%E5%AD%A6%E4%B9%A0/3729729
[6] Artificial intelligence makes automatic driving a reality Xinhuanet client 2018-10-19
https://baijiahao.baidu.com/s?id=1614713922655171357&wfr=spider&for=pc
[7] Deep learning: application of artificial intelligence in automatic driving July 24, 2018
https://baijiahao.baidu.com/s?id=1606707037958720590&wfr=spider&for=pc
[8] New jingche comments on the frequent occurrence of Tesla's automatic driving accidents, sounding the industry safety alarm on December 16, 2019
http://www.bjnews.com.cn/auto/2019/12/16/662511.html
[9] Experience the cutting-edge technology charm of Hikvision from face recognition black technology 2018-06-27
https://www.sohu.com/a/238284495_668694
[10] Using the face recognition algorithm based on deep learning, Hikvision has developed the scheme of close range face access system 2018-10-16
http://www.elecfans.com/d/799038.html
[11] Face painting payment must put safety first October 24, 2019
https://baijiahao.baidu.com/s?id=1648231300830465117&wfr=spider&for=pc