Weighing the Pros & Cons of Facial Recognition

With facial recognition technology often making the news for the wrong reasons, we thought it would be helpful to look closely at the advantages and disadvantages of this technology and other similar AI technological systems. Our hope is that by carefully weighing the pros and cons of this rather new technology, people can make more informed decisions about facial recognition technology, maximizing its benefits and minimizing the limitations of its systems.

Although facial recognition is an advanced and powerful biometric technology, many early technological missteps, especially in the areas of personal privacy, have caused the technology to receive a bad reputation rather early in its development.

Luckily though, during the pandemic, some of the benefits of facial recognition technology shined brightly in fields such as remote service enrollments, no-touch biometric access control, and health and mask monitoring; our hope is that the recognition of benefits like these continue to change how people think of facial recognition technology and that its acceptance among the greater public continues to increase.

The big debate: Is facial recognition considered “good” or “bad”?

There is a lot of debate on whether facial recognition and other related artificial intelligence (AI) technologies are “good” or “bad.” We believe that ultimately, no technology can be immediately considered a good or bad technology; it is all based on the responsibility of the people who make use of it.

Many who argue that facial recognition technology is “bad” will often attribute it to issues of personal freedom and an invasion of privacy, especially in countries with governments and businesses that act more totalitarian and are eager to exert stringent public monitoring controls to preserve their power.

This isn’t to say that facial recognition technology cannot be abused in the free world, but we think that facial recognition is similar to the advent of the internet, along with internet advertising, online media, and of course the rise of social media; we have seen that the newness of these technological advances may seem daunting at first and will face stages where they are largely seen as “bad,” but as time has progressed and the pros of these technologies shined through and became essential parts of many advanced economies, the fears associated with this new technology died down.

We believe that this will hold true for facial recognition technology too.

Five Benefits of Facial Recognition Technology

Facial recognition is one of the most secure and fastest biometric verification solutions on the market. Although there are many benefits a business may enjoy by adding facial recognition to their current workplaces, we think these five are particularly beneficial:

1. Increased safety and security for businesses, employees, and customers
2. Enhanced protection from identity theft
3. Higher accuracy and speed
4. Ease of installation (flexibility and scalability)
5. Improved productivity and customer satisfaction

Now let’s take a closer look at each of the benefits.

1.Increased safety and security for businesses, employees, and customers

For businesses, facial recognition biometrics systems can be used for security access control and surveillance, helping companies protect their valuable assets. Meanwhile, for employees or customers, the monitoring and enforcement of anti-epidemic measures as well as the safeguarding of personal assets in either the real world or digital domains help increase the safety of those involved in businesses using AI facial recognition technology.

2.Enhanced protection from identity theft

Biometric technologies also aid in protecting an individual’s identity. The risk of identity theft is lowered when one uses their face as a main identifier (rather than a government-issued ID that can be stolen, such as a passport, ID card, or driving license).

Fortunately, CyberLink’s FaceMe is especially secure, having achieved a perfect iBeta's Advanced Level 2 Anti-Spoofing Test score, ensuring that 2D fakes and even 3D printed and resin facial masks could not fool the AI-enhanced facial recognition technology.

3.Higher accuracy and speed

Facial recognition technology allows businesses an accurate, smart, secure, and fast means of recognizing individuals in a big crowd. For companies desiring to quickly clock-in employees or grant only authorized personnel access to select buildings, solutions like FaceMe are ideal.

Even the busiest of commercial premises, FaceMe provides a potential 500+ frames per second face matching technology. Moreover, CyberLink’s FaceMe has achieved a 99.81% accuracy rating in industry-recognized NIST FRVT 1:N VISA Border tests, and the technology has a less than one in a million chance of incorrectly authenticating an individual.

4.Ease of installation (flexibility and scalability)

Installing and commissioning an AI-enhanced biometric security infused business solution might sound like a costly endeavor to most, however, FaceMe is actually very easy to implement, use, and scale. CyberLink’s facial recognition solution runs on a wide variety of systems, from devices as common as smartphones, tablets, PCs, and video management systems, to devices as specialized as cameras in hundreds of locations and dedicated GPU-accelerated servers.

FaceMe is compatible with a wide variety of operating systems including Windows, Linux, Android, iOS, and others. If a business has already invested in a video management system (VMS) for security and surveillance, FaceMe can seamlessly integrate with solutions from major providers like Milestone, AXIS, VIVOTEK, Network Optix, and Genetec.

This degree of flexibility and scalability has made it easier than ever for organizations to upgrade to smart surveillance or smart access control without incurring big costs.

5.Improved productivity and customer satisfaction

Facial recognition-based systems cut down on the need for keys, access fobs, and ID cards, all of which can be easily lost or stolen. Using contactless biometric technology like facial recognition engines also creates a more hygienic working environment.

Facial recognition time and attendance systems, for example, cut down the long check-in/out lines during rush hours. The systems also have the benefit of accurately compiling employee attendance records and simplifying payroll.

The banking, financial service, and insurance (BFSI) businesses are another segment that benefits from facial recognition as part of a rigorous eKYC (electronic Know You Customer) system. Thanks to facial recognition technology, industries like these benefit from following anti-money laundering regulations and safeguarding customer investments with more ease than before.

Customers involved in businesses using facial recognition technology benefit from faster, easier, and more secure access to products, locations, and services. All of this helps improve the customer experience and increases customer satisfaction.

Five Common Concerns of Facial Recognition

Although facial recognition offers a powerful and efficient identity verification, we do want to also address the common concerns and hesitations that most businesses face before investing in this technology:

1. Legal and privacy concerns
2. Storage capacity
3. Data vulnerabilities
4. Computational requirements
5. Camera quality and lighting concerns

Let’s look at each of these possible concerns in-depth:

1.Legal and privacy concerns

Some individuals express concern about facial recognition systems storing their data in the form of photos and information about their physical activity. Many data privacy laws and regulations (such as GDPR, CCPA, BIPA, and LGPD), consider biometric information (not just photos) to be personal information and thus have installed strict requirements to safeguard personal privacy.

For example, under these regulatory laws, consumer data gathered by facial recognition systems can only be stored with consent and must be protected and allowed access/deletion rights just like any other form of stored digital data used by businesses today.

In order to stay compliant with these privacy laws, FaceMe facial recognition provides their customers with these critical features:

1. Active Facial Recognition: When a user’s face takes up more than 25% of the screen view, the system will ask if the user agrees to have his face data extracted. FaceMe will only collect, analyze, and store this data when the user nods his head as a form of consent.
2. Right to Erasure (AKA, Right to be Forgotten): FaceMe also provides users with the right to delete their information upon request whenever they desire.
3. Right to Data Portability: FaceMe provides users the right to export their data whenever they desire.

For more information on privacy protection regulations in the United States,please see our white paper Responsibly Using Facial Recognition in the U.S.

2.Storage capacity

Many people recognize that a facial recognition biometrics system in businesses with hundreds of employees often also requires a large amount of storage to make the best use of facial recognition. This is especially true of deployments that combine VMS (video management systems) for security access control with solutions like FaceMe Security or FaceMe TimeClock. Video data, which may cover hundreds of cameras for hundreds of hours, is well-known for its heavy storage requirements.

We find that many of our customers who are able to run their software on AIoT or edge devices, storing their data in the cloud, are able to work around storage issues like these. We encourage businesses interested in potentially using FaceMe technology to evaluate their options and review our business use case for facial recognition.

The viability of using facial recognition with cloud storage will ultimately depend on factors like camera coverage, the flow of people through the business, etc.

3.Data vulnerabilities

Individuals have realized that digital technologies are naturally susceptible to data vulnerabilities from hackers and identity thieves. Although this is a reasonable concern, steps can be taken to significantly reduce the risk of such attacks.

We encourage our customers to always keep their company data protected by all reasonable means, taking necessary procedures such as keeping their software and underlying systems up to date, and regularly safeguarding their commitment to user access policies.

4.Computational requirements

Many businesses worry that facial recognition technology has large computational requirements to run such advanced software. Although it is true that the businesses that invest in more capable computer systems will be best suited to reap the benefits of faster processing, we also recognize that this is not feasible for all businesses, which is why FaceMe is designed to be scalable.

In many cases, customers can still use FaceMe just with a compact smart device, and the use of a connected or AIoT/IoT device with a built-in or attached camera is enough to run FaceMe.

There are many deployment options facilitated by FaceMe SDK for business needs of all sizes. Developers are even able to enable GPU acceleration, utilizing OpenVino, NVIDIA Jetson, ARM, etc. to speed up deep-learning algorithms. Check our FaceMe SDK System Requirements for more information.

5.Camera quality and lighting concerns

AI-enhanced biometric technology is highly accessible due to the availability of smart devices with cameras, which can be easily utilized to carry out facial recognition. Of course, facial recognition works best with high-quality cameras and ample lighting, especially those offering 3D, ToF, and IR capture options for high-traffic facial recognition use-cases and ideal lighting situations so cameras can operate at faster frame rates, and for capturing clearer detail to feed the AI; however, this is not always necessary.

CyberLink’s FaceMe is designed to work with a wide array of cameras, from the entry-level 2D smart device or webcam to high-end multi-lens 3D capture devices. It works accurately with the full array of camera systems, with 100% anti-spoofing for 2D and 3D cameras.

Legal considerations when using facial recognition technology

There has been a growing concern regarding facial recognition systems storing data, both in the form of photos and recorded movements. Although it is true that cameras facilitating facial recognition have the potential to capture and store such data on associated computer equipment, companies using such technologies are still obligated to comply with certain rules and regulations that help safeguard individual privacy and allow users to request data deletion.

Some of these data privacy laws and regulations that biometric technology companies must adhere to include GDPR, CCPA, BIPA, and LGPD, as these guidelines consider biometric information to be personal information. Due to these institutions and their laws, facial recognition vectors fall under these regulations, as would fingerprints or iris scans. Indeed, these issues are complex and regionally dependent, but they are nothing that companies considering facial recognition technology cannot handle.

We stress the importance of taking the time to consider which type of facial recognition is best for the company, receiving user consent where legally required, and being rigorous in protecting such data and applying security software updates.

Our Responsibly Using Facial Recognition in the U.S. white paper talks about how businesses in different states can lawfully use facial recognition to enhance security and productivity without violating users’ privacy. The white paper covers the following:

• The different types of facial recognition and how they work
• Why the handling of consent, data, and retention is key
• A summary of existing laws at both the federal and state levels
• How easy it is to use facial recognition responsibly and legally

FAQ