Facial Recognition Technology: Challenges and Use Cases

Although facial recognition is widely used all over the world, it has received a lot of backlash from proponents of privacy and equality.

Privacy Concerns

Advocates of protecting privacy feel that face recognition technology is all set to deprive them of their constitutional right. After all, not everyone fancies being recorded and identified while jogging at a public park or attending a sports event. Despite these privacy concerns, video surveillance is essential in public places as it helps ensure physical security. So, to strike a balance between the demand for personal privacy and providing physical security, solutions like Scylla were designed. Scylla’s Face Recognition System does not store the data but provides the client with an option to do so on their local network defining the storage period and data access in accordance with their security infrastructure and data security guidelines.

Racial Bias

You may have come across the term racially biased face recognition applications and the criticism surrounding it. This happens when the face recognition module is focused on those belonging to a particular ethnicity rather than on individuals from diverse backgrounds. We discussed it in further detail in our article by Dr. Armen Ghambaryan “How to ensure accuracy and sensitivity of the face recognition system”.

So, one needs to look for such face detection and recognition solutions that makeuse of state-of-the-art deep learning techniques and use balanced database to train their system. At Scylla, we consciously made product decisions that are in line with our ethics and mission. Therefore, Scylla team has deliberately built ethnically and gender balanced datasets in order to eliminate bias in face recognition.

How is Face Recognition used for Physical Security?

Face recognition technology and AI video analytics played a pivotal role in video surveillance and made it far more accurate and proactive compared to manual surveillance. This is the very reason why the FBI, Interpol, and other law enforcement agencies have been using it for so long. Although there were some concerns regarding the accuracy levels of the NGI-IPS algorithm, it now seems to be resolved.

As such, the FBI now engages vendors that offer advanced solutions with an accuracy level ranging between 99.12% to 99.72%. However, the use of face recognition in physical security is not limited to law enforcement agencies. Large businesses, airports, manufacturers, educational establishments and sports facilities increasingly adopt it to bolster their physical security as well.

Some of the most common use cases of applying face recognition technology for physical security:

Missing People Detection

According to the report released by the US Department of State, 25 million people are being compelled to do something they’d rather not do. This includes forced labor, slavery, drug dealing, and other illegal activities.

Despite all the technological advancement we talk about, every year 600 to 800 thousand victims are trafficked. As these victims are usually reported missing, those data can be used by face detection systems to trace and rehabilitate such victims.

Also, it can help in the detection of missing individuals who cannot protect themselves or return home on their own. For example, children, mentally disabled individuals, physically disabled, or those who have succumbed to drug abuse.

Access Control

A combination of facial recognition software and a local access control system can provide a contactless access control experience which is the need of the hour.

At Scylla, we have created an all-in-one next-generation Workplace Safety Solution that makes use of this and other AI-powered Scylla physical security systems to enable a far safer and more secure office environment.

Attendance Control

Time and attendance management can either be a tedious task or one that takes just a few seconds. The outcome depends on how you choose to manage the entire attendance system. Nowadays you could do that with a single face recognition-based attendance management system that updates the HR and payroll records in real-time.

Unfortunately, many businesses continue to follow the archaic systems that follow a two-fold approach — an application that performs the authentication only and another one where those data are manually entered and stored. This traditional approach brings forth two problems — increased costs and time lag in the availability of updated data. Managers need real-time data to optimize their resources and any delay can create roadblocks. Businesses can easily overcome these hurdles by using a face recognition-based attendance system that is well integrated into the HR and payroll modules.