Since the beginning of the century we have witnessed at least three epidemics of transboundary nature which have significantly influenced the public perception of diseases as such. Severe Acute Respiratory Syndrome (SARS), which spread across the world in 2002-2004, outbreaks of avian (“bird flu”) and swine (“swine flu”) influenza viruses in 2003-2015, and the Middle East Respiratory Syndrome (MERS), which emerged in 2012-2013 in Saudi Arabia and in 2015 jumped to South Korea. The outbreak of the Ebola epidemic in Africa somewhat stands alone – its difference from all the above forms of viruses is that Ebola does not spread through the airborne route which has facilitated its successful containment.
It is safe to say that technological progress is, in some way, the main contributor to the global spread of infections. The globalization of technology has advanced movement of people over long distances, be it car, bus or train trips from one city to another, or long cross-border flights. The speed with which viruses can be spread around in our actual living conditions is astonishing. The interesting thing is that the information about the emergence of the virus gets to us even quicker, thanks to the development of information technologies.
Today we are witnessing the growing spread of the decade’s forth epidemic, and its largest one – on several indicators: in the number of cases, the progress of the virus, and the level of information support. Thus, the traffic on the Chinese website which publishes information on the distribution of coronavirus 2019-nCoV has reached 4 million views per hour, with the total number of visits exceeding one billion since its launch.
So which attribute do all these epidemics have in common? That’s right, face masks. Given the threat of the virus spread, seeing someone wearing a medical mask or even a respirator from an extraordinary phenomenon turns into a normal thing, or even a necessity.
Here is a picture from 2003, for example – “Ladies ballet class in Hong Kong”
Young ballet students in Hong Kong wear protective masks in April 2003, during the height of the SARS scare in the city. Hong Kong’s lessons from that outbreak have in some ways helped the global response to Ebola. (Vincent Yu / Associated Press)
A typical day out in Delhi during the swine flu epidemic, 2009. Source: Reuters
People wearing masks wait in a queue for a H1N1 flu screening at a hospital in New Delhi August 11, 2009. REUTERS/Fayaz Kabli.
In today’s China, “face masks” went even further.
Wearing a face mask becomes perfectly normal in response to a viral respiratory infection epidemic. In fact, with a risk of becoming infected anytime, anyplace – be it in a public transport, in the theater, in the store or at a dinner party – a face mask is gradually moving into the category of life-sustaining items, such as clothes. Leaving the house in the midst of an epidemic without a face mask can be translated as bad form. In contrast, wearing a face mask is seen as a sign of self-preservation and concern for the health and safety of others around you. And we are not aiming at covering our faces here in doing so – that’s just a side effect. Well, sort of.
Does not this make you wonder what might the future of recognition technologies look like, when everybody is wearing a mask that covers a large part of their faces? In the “Internet of everything” era, face recognition systems become if not the core, but a very important element of public safety, target marketing, and social interactions in general. They are now integral parts of city infrastructure — security cameras are hidden in the areas where you won’t even suspect them to be, such as: streets, airports, and grocery stores, hotels, or sports stadiums. They are there not only to prevent crime by tracking unsavory characters, but also to contribute to effectiveness of targeted sales and advertising, through “knowing-your-customer”. It seems like KYC has taken somewhat more literal meaning.
For most face recognition algorithms to work, the person has to be barefaced, because those algorithms evaluate the face “as a whole”. We are not talking about iris recognition here, which is an automated method of biometric identification used as part of building access control systems. When the face recognition system spots a medical mask or respirator on the human face, it will most likely respond with a “face is covered”, and refuse further recognition attempts. The chance of recognition failure is even higher in cases where there are some decorative elements on the mask, such as prints or texture: could be an abstract pattern, or a jaw, or Darth Vader’s helmet. Clearly, the scenario where every passerby in the city wears a face mask is not anticipated by the developers during systems testing, the described implications therefore aren’t prevented. However, such a scenario does not seem to sound that surprising anymore.
Face recognition systems will quite obviously become useless in case the majority of people start protecting themselves with a face mask — unless we see some major improvements in the algorithms of those systems, of course. Not only that: relying on them could be dangerous, since the percentage of false positives might increase sharply, which will inevitably lead to malfunctions.
Recognition technologies will certainly be further developed, algorithms will change, and it is conceivable that we will soon witness the emergence of a solution that will be able to provide recognition results at a fairly acceptable standard with only uncovered part of the face. Those results could be then taken into account in the decision making. As of now, the flow of information on the coronavirus epidemic is not only leading to the collapse of cross-border online trade, but it has also significantly affected those sectors which have just started getting accustomed to face recognition systems, and the development of the face recognition technology itself.
Green AI-powered scanner SDK of ID cards, passports, driver’s licenses, residence permits, visas, and other ids, more than 1810+ types in total. Provides eco-friendly, fast and precise scanning SDK for a smartphone, web, desktop or server, works fully autonomously. Extracts data from photos and scans, as well as in the video stream from a smartphone or web camera, is robust to capturing conditions. No data transfer — ID scanning is performed on-device and on-premise.
Automatic scanning of machine-readable zones (MRZ); all types of credit cards: embossed, indent-printed, and flat-printed; barcodes: PDF417, QR code, AZTEC, DataMatrix, and others on the fly by a smartphone’s camera. Provides high-quality MRZ, barcode, and credit card scanning in mobile applications on-device regardless of lighting conditions. Supports card scanning of 21 payment systems.
Automatic data extraction from business and legal documents: KYC/AML questionnaires, applications, tests, etc, administrative papers (accounting documents, corporate reports, business forms, and government forms - financial statements, insurance policies, etc). High-quality Green AI-powered OCR on scans and photographs taken in real conditions. Total security: only on-premise installation. Automatically scans document data in 2 seconds on a modern smartphone.
Green AI for Tomographic reconstruction and visualization. Algorithmization of the image reconstruction process directly during the X-ray tomographic scanning process. We aspire to reduce the radiation dose received during the exposure by finding the optimal termination point of scanning.
Please fill out the form to get more information about the products,
pricing and trial SDK for Android, iOS, Linux, Windows.