Facial recognition technology and the paradigm shift it has brought about
Until recently, the idea of technological developments that could differentiate between human faces was the matter of an H.G Wells novel. Today, from glancing at an iPhone X to unlock it, to those few seconds of staring at a camera during immigration at most airports, facial recognition has almost seamlessly been integrated into our daily lives.
From when it was first developed by American mathematician Woodrow Wilson Bledsoe using a RAND tablet in the 1960s, to when Facebook began using face ID in 2010, facial recognition technology has been used for a variety of purposes, and its applications are being broadened every day. Over the years, companies have pared down many basic inaccuracies in the technology, and despite still retaining many serious issues including racial bias, it has proven to be of immense benefit to society at large.
Today, national security relies on facial recognition to a large extent. A significant number of international airports, such as Roissy Charles de Gaulle airport in Paris, make use of biometric authentication in the form of facial recognition.
According to the 9/11 Commission Report, a system to record the faces of foreign nationals who entered and exited the country was integral to national security, which is why all non-U.S. citizens traveling to the country have their picture taken on arrival. Biometric authentication by face is tougher to bypass inconspicuously, reducing the likelihood of terrorist activities and strengthening national security.
In addition to benefiting national security, facial recognition is used to ensure safety on a local level as well. For the Tokyo 2020 Olympic Games (postponed to September 2021), facial recognition technology was due to be employed to grant access to authorized personnel and improve safety at the event. In India, facial recognition is used for ‘Aadhar verification’, a system of official identification used nationwide. This is similar to measures taken up by other countries, such as China, where its social credit system is supplemented by widespread video surveillance.
Facial recognition technology has created ripples in the world of healthcare too. Face2Gene is an app used by clinicians to detect rare genetic disorders such as Cornelia de Lange syndrome and Mowat Wilson Syndrome. The app allows geneticists to be more certain of their decisions when recommending genetic tests, and this can lead to financial savings for the patients as well.
Facial recognition is also used for emotion detection in patients, which allows healthcare facilities to tweak their services according to the patient’s preferences. In addition to the above, there are a number of healthtech projects related to this technology which are currently underway, such as a test being developed by RightEye to detect autism spectrum disorder. These, if successful, may significantly change the way mental health is assessed and diagnosed.
Besides security and healthcare, facial recognition is also used in places of worship; a number of churches use facial recognition technology to track the attendance of its members. Face-Six LLC, a face recognition software vendor, has a division called Churchix, which eliminates the need for manually taking attendance at churches. Moshe Greenshpan, CEO of Face-Six, says that the company has worked with over 200 churches so far.
As with most developments, the technology of facial recognition is not airtight. With the rise of each innovator, there is a hacker, persistently seeking out the Achilles heel of facial recognition technology. For instance, Adam Harvey, a designer and entrepreneur, developed ‘CV Dazzle’, which uses makeup and hairstyling to get around the frameworks used by open-source computer vision library OpenCV.
In May 2018, Forbes published an article about how researchers at the University of Toronto developed an algorithm to disrupt facial recognition technology. In a sample of 600 faces from a wide range of ethnicities, lighting conditions and environments, the algorithm was able to reduce the number of detectable faces from 100% to 0.5%.
More recently, during anti-extradition law protests in Hong Kong in 2019, protesters used laser pointers and masks to avoid being recognized by facial recognition technology used by the police.
These attempts to subvert the often-eerie recognition capabilities of artificial intelligence mostly stem from fears regarding data privacy and security. According to a May 2018 report, the FBI can access 412 million faces for their research. This is particularly worrisome considering the high probability of inaccuracies in these technologies, especially for minority groups.
The National Institute of Standards and Technology conducted a test that showed that Idemia, a software used extensively by the FBI, exhibited inaccuracies in recognizing faces. The software was 10 times more likely to misidentify the face of an African-American woman, than the face of a white woman. This can exacerbate the issue of racism in the society and lead to exploitation of these technologies.
Facial recognition technology is certainly a hallmark of man’s innovation and prowess, and has the potential to revolutionize the way we function in multiple fields. However, a number of governments have acknowledged the downfalls of this technology and have taken measures to ban it, or at least restrict its use.
Following protests against the unjust killing of African-American George Floyd in May 2020, IBM stopped offering its facial recognition product for mass surveillance or racial profiling. Between June, July and August 2019, many states in the U.S., including San Francisco, banned their police forces from using facial recognition technology.
Despite its darker side, it is undeniable that facial recognition technology has the potential to facilitate landmark developments in different fields. Lawmakers must weigh the advantages and disadvantages associated with unrestricted use of facial recognition and pass well-thought out laws regarding its role in society, in order to fully utilize and optimize the benefits it posits.