Processing...

Unisol International 8024 NW 90th St.
Miami, FL 33166, USA.
Tel. 1-305-885-2656 Fax. 1-305-489-2685

By Alberto Jessurun / April 14, 2021 / Blog

What Hotels Should Know About Face Masks and Facial Recognition Technology

Both security and health are paramount for hotel operators. How can facial recognition adapt to accommodate mask compliance during Covid-19?

On a typical pre-pandemic Halloween night, hotels along the Las Vegas Strip made it clear that revelers could not enter the premises while wearing masks. Masks presented a security issue, so guests were required to remove them (or plan costumes accordingly) if they wanted to step inside.

Amidst the pandemic, masks are now required inside these same hotels. This begs the question: As the Covid-19 crisis continues and face masks are increasingly required to engage in day-to-day public life, how can hotel security adapt?

As of March 10, 34 states in the United States require face masks to be worn in public. Even among the states that do not have mask mandates, face masks are often still strongly encouraged or required in certain industries and cities.

For hotel operators, this presents a conundrum: Guest and employee security remains just as important as it was pre-pandemic, but public health and local regulations are important considerations.

According to experts, Americans may be wearing face masks through 2022, potentially even months after day-to-day life has returned to normalcy otherwise. With this in mind, hotel operators must plan their security measures to accommodate mask compliance on a long-term scale. Prior to Covid-19, facial recognition technology was a popular way for hotels to improve their security. Face masks complicate this option, but they have not made it impossible.

How does facial recognition work?

Facial recognition technology continues to innovate, and it is on its way to being adopted by even more industries for even more applications. Facial recognition works by analyzing key facial features — such as the eyes, nose, and mouth — and comparing them to a single saved a face or faces in a database.

The technology either verifies the face is a match, presents a series of matches, or comes up with nothing. Advanced facial recognition technology can go above and beyond this, identifying other characteristics in addition to facial features, such as age, gender, and even emotions based revealed by facial expressions.

Hotels can use this technology to identify prohibited individuals, find the culprit after a crime, or even prevent crimes by flagging suspicious behavior.

Face masks and facial recognition

Face masks serve the important purpose of preventing the spread of Covid-19. Unfortunately, they can also make it difficult for facial recognition technology to do its job. In a recent study about face masks and facial recognition, tests showed that masks can cause an error rate of 5% to 50%.

Interestingly, the color of face masks played a role in this, with darker masks raising the rate of error. The shape of face masks studied also played a role as rounded masks resulted in fewer errors.

These tests used approximately 6 million photos to reach the conclusion that face mask usage raises facial recognition technology error rates substantially. In an unmasked setting, error rates are about 0.3%.

One-to-one matching vs. one-to-many

Worth noting about the study is the fact that the tests used a one-to-one matching system. A one-to-one system matches one face against a single saved set of data to determine a match. This type of facial recognition is commonly used to unlock smartphones. One-to-one matching can be affected by lighting, camera quality, and positioning of both the subject and the camera.

The aforementioned study did not test facial recognition technology in a one-to-many setting. Unlike one-to-one, one-to-many technology matches against a large database of faces. The one-to-one matching tests suggest that one-to-many matching (which is used by hotel operators or other facilities to monitor large groups of people) may face an even more substantial error rate due to the inherent challenges of picking out a face in a crowd with varied lighting, angles, obstructions, etc.

How hotel security can adapt to face masks

Given that face masks will likely be around for the foreseeable future, it is important for hotels and anyone else using facial recognition technology to adapt. Fortunately, technology that accounts for mask compliance is on the way. In the meantime, hotels would be wise to take their cues from industries and locations that have faced these challenges pre-Covid. For example, hospitals where face masks have long been the norm, or religious settings where facial coverings are worn.

Hotels also might consider offering complimentary face masks in a light color to combat the error rate uptick with dark face masks. Camera position and lighting are also crucial to combat this. Hotels should also have backup security measures in place, such as key cards that restrict access to certain areas, sign-in sheets, and enhanced staff training.

Finally, consider facial recognition technology that relies on an algorithm targeting the eyes and upper facial features. This technique will likely become prevalent as mask-wearing continues.

Technology from Unisol

For the most modern, best-in-class security solutions, contact Unisol International. We offer a robust product catalog, ship worldwide, and have the in-house expertise to help solve your toughest IT challenges.

Working on a Project?

We'll help you find the perfect product and service package for your needs. Speak to a specialist today for immediate assistance.

305-885-2656 Hablamos Español

What Hotels Should Know About Face Masks and Facial Recognition Technology

Both security and health are paramount for hotel operators. How can facial recognition adapt to accommodate mask compliance during Covid-19?

How does facial recognition work?

Face masks and facial recognition

One-to-one matching vs. one-to-many

How hotel security can adapt to face masks

Technology from Unisol

recent posts

Unlocking the Power of a Unified Retail Strategy: Integrating POS with E-commerce

Balancing Security and Privacy With Biometric Systems

Beyond Simple Counters: An Overview of Mobile POS Systems and the Future of Retail

categories

Working on a Project?