Social acceptance and technological maturity in the West
From the order of science fiction at the beginning of the millennium, considered unreliable, facial recognition has since made immense progress, thus following in the wake of those already made by artificial intelligence. One could even speak of a revolution: in June 2015 Google (with FaceNet) thus obtains, in the benchmark test Labeled Facebooks in The Wilda score – unprecedented for the time – of 99.63% accuracy, almost identical to human performance.
Operating since the mid-2010s, the use of facial recognition has however long been slowed down in the Western world by the persistence within it of a strong libertarian tropism. An obstacle that has not been experienced in Asia, and in particular China, for which the concepts of harmony and order traditionally supplant that of individual freedom. This is one of the reasons that explain the American delays in this area, compared to the Chinese. But thanks to the pandemic, facial recognition, like all so-called “societal control” technologies, is now better accepted by the populations of the West. Necessity being law, even the most refractory take it up: an immense majority of French people have, for example, approved the vaccination pass. Like all other industrial applications of artificial intelligence, facial recognition could thus, in the near future, be integrated into the daily lives of Westerners. Especially since, in France as in the United States, the regulation of the sector promises a strict framework in the face of certain legitimate fears of public opinion, such as respect for privacy. And that there is – not yet – question of any social credit in the Chinese way.
Facial recognition: Meta indicted by the State of Texas
Facial recognition and security
For Thales, the French electronics giant, facial recognition would indeed offer efficient solutions to two complementary security requirements, namely identification and authentication. According to the French industrialist, the first is to answer the question of the identity of an individual, while the second aims to verify its veracity. The solution facial recognition system (LFIS) would thus have a face acquisition rate in less than 5 seconds of 99.44% and an actual identification rate in less than 5 seconds of 98%, according to the Homeland Security Science and Technology Directorate the United States. The GAFAMs are obviously not left out in the field, since they have implemented facial recognition in their products to – among other things – improve the security of iPhone Xs, or even in Google Photos to sort the photos automatically, according to faces. recognised.
But facial recognition technology doesn’t just identify or authenticate a face, it can also be integrated into comprehensive security solutions through ‘computer vision’, which involves empowering systems to “see” and identify objects. Remark Holding, a pure American player listed on the Nasdaq and very well established in China via its subsidiary KanKan, has thus been able to develop the necessary technology in this segment, in addition to its more traditional positioning on facial recognition, to supply since 2021 to Brightline , a Florida railway company that has experienced significant security problems in the past, a solution for detecting unauthorized intrusions into its trains, “while providing real-time, actionable predictive analysis from data collected while monitoring long stretches of track and yard areas for unusual behavior” indicates a press release from the American company. Moreover, the applications of facial recognition are not limited to the safety/security sector. The applications are many, especially in health and customer experience.
Numerous applications in the field of health
Thanks to deep learning and facial recognition of emotions, it is now possible to monitor a patient’s medication consumption, support pain management and even detect rare genetic diseases, for example DiGeorge’s syndrome, and what with sharpness 96.6%. More prosaically, facial recognition has also been widely used to detect high temperatures during the pandemic.
The latest notable development in facial recognition is unexpected to say the least, but quite promising for commerce and customer experience. The latter could indeed greatly benefit from the installation of cameras in points of sale in order to analyze customer behavior and improve their journey. It is once again in China that the revolution is underway: KFC and Alibaba have been leading a test facial recognition payment system in the city of Hangzhou. The Middle Kingdom is certainly at the forefront of progress, already in 2015, Jack Ma, the boss of Alibaba, had unveiled a selfie payment system at the Cebit fair in Hanover based on the technology Face++ from the Chinese start-up Megvii.
Market with high growth potential in North America
Long before the pandemic and when mistrust was at its peak, a study of June 2019 from the firm Markets and Markets nevertheless estimated the potential turnover of the facial recognition market at 7 billion dollars in 2025, with an average annual growth rate of 16%.
According to the firm, if “the services segment has the highest average annual growth rate during the forecast period”however, it is the law enforcement sector that will have “largest market share”, while most of the revenue will be generated in North America. Especially since the United States is on track to accelerate the adoption of artificial intelligence with the current infrastructure bill, which promises increased funding for the use of artificial intelligence technologies for American public and private actors.
It is therefore the major players in this region that should do well, such as Animetrics, Remark Holding, Amazon or Microsoft. But the competition promises to be fierce: it will be necessary to count on competition from European players such as the French Thales or the German Cognitec Systems, without forgetting the Japanese from NEC Corporation or the Israelis from Id3 Technologies.