In our introductory guide to facial recognition, we provided a full and comprehensive overview of the technology and how it works. To recap, facial recognition is a biometric technology that identifies facial vectors and features and matches them to a pre-enrolled individual. This technology is best used across an edge computing infrastructure. The technology has been around for several years and is now being used on a large scale. In this article, we are leveraging our latest insights about facial recognition’s use cases and trends, based on our unique perspective, as the developers of the top tier FaceMe® engine and solutions.
Today, many consumers all over the world purposefully interact with facial recognition technology on a regular basis. Chief among use cases is to unlock a mobile phone. This is one of the most ubiquitous use cases of facial recognition technology. There are a whole lot more uses of facial recognition out there, several of which are widely adopted already.
Facial recognition technologies, such as CyberLink’s FaceMe®, offer many capabilities which are already being adopted or can be deployed to add value across many vertical industries.
We will now explore specific use cases for facial recognition technology across industries and highlight considerations for implementation. We will also touch on the innovation that is driving the development of solutions impacting growth in key sectors of the global economy.
Access control is the selective restriction of access protocol for individuals to a specific place or resource. To demonstrate how facial recognition can enhance access control, let’s consider the following examples.
Facial recognition is extremely effective to automate and enhance personnel facility access processes, without sacrificing security.
For example, a commercial building rented out to various organizations can use facial recognition technology to ensure safe and secure access to the facilities. It can restrict access to authorized employees or pre-registered guests through face detection and recognition. Even for those who are granted access, it can guide, limit and monitor where people are allowed to go.
Smart locks are growing in popularity to provide access to homes, restricted buildings, rooms, devices and cabinets. They work by granting access using a wireless protocol and cryptographic key for authorization.
Facial recognition can be embedded into smart locks to provide credential access to restricted areas for authorized people only. They are used in both commercial and residential settings. Only once an individual’s identity has been recognized and authenticated through facial recognition, will the smart lock grant access.
Passenger boarding is one of the many bottlenecks of air travel. No matter how inventive airlines can get at sequencing passengers in groups and reminding them of the process, scanning boarding passes or mobile phones slows everything down and introduces multiple points of failure.
Airlines all around the world have started implementing self-boarding turnstiles using facial recognition. Since TSA already has passengers’ photo ID data, matching live face capture at the gate with the information on file is an easy step that makes the process quick and safe, while virtually eliminating the risk of boarding error.
Custom and immigration control upon entering a country is one of the least enjoyable, yet necessary experience for any traveler. In recent years, several countries have been deploying self-service kiosks that let people pre-enter declaration data while waiting in line, reducing the time spent with immigration officers.
A growing portion of these kiosks are now powered with facial recognition. ID scanned at the kiosk is matched with the live face capture to validate identity and populate the declaration forms with information on file. Registered frequent travelers, are identified by the kiosk and benefit from a streamlined, often fully automated process. US travelers are familiar with the Global Entry kiosks that have quickly become a sought after privilege.
Elevators not only carry people to upper floors in seconds, but can be used to control access to restricted areas of multilevel buildings, by requiring the use of access cards for example.
Facial recognition can make the process much safer and simpler. Consider a luxury penthouse atop a new development that just opened downtown, or a research and development center on the 70th floor that only those with proper clearance can access. With facial recognition technology, access to these floors is granted or denied as soon as an individual steps in the elevator car. And automated alerts can be sent if someone gets out on an unauthorized floor.
At pharmacies and hospitals, stockpiles of medicine are closely monitored with limited access across personnel for security, safety and regulatory purposes. To ensure the highest level of security, facial recognition can authorize access to approved staff for high-value medicine and other medical resources, send alerts in case of unauthorized access attempts and keep a complete log of who took which items.
From research facilities, hospitals, factories and warehouses to agriculture and mining, there is a plethora of specialized equipment and machinery that requires strict access and operation control, tracking and reporting. Traditional control systems vary greatly in effectiveness and sophistication, ranging from keys and paper log books to keycards and computerized tracking. Embedding facial recognition access control in the machinery enables precise contactless face login and detailed usage log, with alerts for issues such as unauthorized login attempts, operation outside of allowed hours or exceeding safe time limits.
Know Your Customer (KYC) is a global practice in financial services regulated by law. It requires financial institutions to make a reasonable effort to verify the identity, suitability and risks of their business relationships, including their clients.
The process typically involves verifying customers’ identity by reviewing and validating state-issued documentation, such as a driver’s license or passport, and cross-referencing with background documents from different sources to combat money laundering schemes for banks, financial services firms and insurance companies. KYC practices also help prevent fraud and identity theft, and allow access control to products and services restricted by law. For example, validating a customer’s age to purchase cigarettes or alcohol from an automated vending machine.
Facial recognition can transform traditional KYC processes, making them more robust and efficient. It also unleashes a wide array of online financial services by digitizing the entire KYC process and enabling customers to perform the entire process themselves, from anywhere, using any PC or mobile device. Adding facial recognition to the process results in electronic Know Your Customer (eKYC) verification. eKYC is a remote and fully digital process using facial recognition to match someone’s live face capture to an official ID scanned or already on file, and confirm the person’s identity prior to granting access to services and products.
The process typically works like this:
Scan photo ID
Capture a live photo
1:1 face match comparison
Here are three examples, among many, illustrating the use of facial recognition for eKYC.
Surveillance systems are omnipresent across the world in residential, commercial and public settings, contributing to societal wellbeing and the general safety of everyone.
From Wi-Fi connected home cameras to sophisticated CCTV systems, video surveillance systems typically all offer effective real-time monitoring and can keep recordings of past event for review and analysis. Typically, their effectiveness is dependent on individuals watching real-time video feeds or recordings, and then manually alerting designated personnel of any event. This process is cumbersome, expensive and unreliable.
Facial recognition can dramatically enhance the effectiveness of security and surveillance systems across sectors and cut the overall costs. It can identify when individuals are in a camera’s sight and recognize anyone who is in the system’s database, automatically sending alerts for targeted human interventions. When applied ethically and constructively, facial recognition can greatly contribute to a safe environment.
Let’s consider the following examples to put the technology into perspective.
Warehouses and factories across the world play a critical role in developed and emerging economies. They are valuable to not only manufacture consumer goods, but to store and ship them worldwide across a vast and complex supply chain network.
Security and safety typically represent significant investments, to protect valuable contents, monitor complex and expensive equipment, and ensure people’s safety from the many hazards inherent to these facilities. Systems equipped with facial recognition can ensure only authorized personnel are accessing restricted areas, or only workers with the proper licenses and credentials can access and operate machinery.
If something was to go awry in a warehouse or factory floor, a smart security system could automatically detect the incident, identify the employees involved and immediately notify people in charge. This allows the people involved in the accident to secure medical attention as soon as possible. There is also the added benefit of the medical team already knowing the people involved and their medical history before even arriving to the scene.
Facial recognition operates by matching live images against a database of pre-enrolled identities, enabling much better monitoring of protected premises and alerting key personnel if an unidentified or banned individual enters.
For example, consider schools where children’s safety is a top priority. School safety has become a key issue for parents and teachers. Smart security systems using facial recognition technology can help authorities better monitor people coming and going throughout entire day. The moment a known criminal or an unidentified adult approaches or enters the school, the system can alert the proper authorities for immediate action.
Similar security features apply to retailers. A retail store might have a history with a customer that repeatedly threatens or coerces the staff, or has been caught shoplifting. A smart security system using facial recognition can be configured to alert in-store security and notify authorities should this frequent shoplifter or disgruntled individual enter the store.
The use of facial recognition and smart security for residential protection is perhaps one of the more exciting innovations. It’s a consumer-facing facet of the technology, operating within the privacy of people’s homes to protect their family and belongings.
A smart home security system equipped with facial recognition can learn to arm or disarm itself automatically when members of the home are gone or present. If the system picks up an individual recognized as a household member, it knows there is no need to raise the alarm. If the armed system identifies an individual on the premises who is not a member of the household or an authorized visitor, it can immediately flag the intruder to both household members and first responders.
For many people, facial recognition is only associated to monitoring, security and safety. The use cases go far beyond and bring about many benefits to personalize retail or hospitality experiences. It can transform the customer experience for loyalty program members who opt-in and register their face picture. It can also provide some level of personalization purely based on key demographics of anonymous visitors. It is a relatively easy task to embed a well-designed facial recognition SDK, such as FaceMe®, in smart signs, interactive kiosks and more. A wide range of businesses like retailers, hospitality operators, restaurants and more can greatly benefit from offering such personalized customer experiences.
Let’s explore some of these use cases further.
Retailers that are looking to improve guest loyalty would be wise to leverage facial recognition. Devices installed in store can be programmed to recognize an opt-in VIP customer and alert staff upon their arrival for a personal greeting and special attention. Facial recognition can also be added to POS systems, enabling opt-in loyalty members to instantly login to their account and safely make a contactless payment, much like how Apple Pay works on the latest iPhones.
At a hotel, facial recognition can be used to identify a top-tier loyalty member the moment they walk through the door. An alert can notify staff so they can provide special assistance and customize their guest’s stay based on known preferences, such as offering a quiet room on the top floor, extra pillows or even a pre-set room temperature.
For restaurants, opted-in diners can instantly be recognized as VIPs and enjoy extra perks and service.
Digital signs have become ubiquitous in recent years, as a tool to deliver rich media in public places, malls, restaurants and retail stores. Adding a camera with facial recognition capabilities to the displays enables the delivery of a personalized, interactive in-store experience. Opt-in shoppers can see personalized recommendations based on their past purchases or stated preferences. Even for anonymous shoppers, a digital sign powered by facial recognition can display more relevant information based on the individual’s physical characteristics, such as gender, age and mood.
For hotels, an interactive sign equipped with facial recognition technology can show opt-in guests directions to their room or offer personalized recommendations for local activities, based on information it already knows about the guest and their preferences.
Adding facial recognition technology to POS devices and self-checkout kiosks can provide a very secure means of cardless, entirely contactless payment for opt-in customers. For large transactions, it can be used for dual authentication in conjunction with a credit card or official ID. It can be implemented anywhere, from the employee cafeteria to fast-food restaurants, and from the grocery store to any retailer. Businesses offering loyalty programs can deploy this technology by implementing a facial recognition SDK to their POS system and by inviting their members to opt-in with their face picture. Global credit card and transaction infrastructure providers are piloting projects and Apple Pay is already providing a solution using iPhones.
Some people might remember the old cigarette vending machines that have all but disappeared due to laws preventing minors from buying tobacco. Adding facial recognition to modern vending machines solves the problem of selling controlled products like cigarettes and alcohol. When prompted, individuals simply need to scan a valid photo ID in the built-in card reader, then look at the camera for a live face capture. The facial recognition engine performs a 1:1 match and enables the purchase.
Facial recognition can also be leveraged for data analysis. In a retail environment, a device with facial recognition can capture customer behavior and demographic data. It can determine if customers look confused or lost in certain aisles, or if they are more inclined to buy a product when it’s placed next to a mirror. These patterns are captured by facial recognition and AI technology, informing the retailer so they can take actions that produce more positive experiences.
Since the COVID-19 pandemic hit in early 2020, the health and safety of individuals in public and private spaces has been a priority, resulting in a series of measures taken around the world, including mandatory mask wearing outside of home. This makes a fascinating use case for facial recognition to keep individuals healthy and ensure that masks are being worn properly, especially in public spaces. You can often find facial recognition embedded into health kiosks, such as the FaceMe® with the Sentry Health Kiosks.
Kiosks equipped with facial recognition use cameras to detect if an individual is wearing a mask, and if they are wearing it properly covering the entire nose and mouth. They can also be equipped with thermal imaging cameras to read temperatures. This helps ensure no one with a fever enters a property. After the individual passes the health checks, they can be granted access. If they do not pass the health checks, designated personnel can be notified to take appropriate action.
The health and safety use cases enabled by facial recognition are boundless. Health kiosks can be useful for retailers, restaurants, hotels, workplaces and offices, public transportation and more. Public transportation is a sector where people can expect to continue seeing health measures taken seriously even after the pandemic, and facial recognition can help greatly.
The pandemic has exponentially increased the health and safety challenges for public transportation operators, especially in areas seeing large volumes of transient people like airports and train stations. Facial recognition helps take on these challenges. Cameras strategically placed at entrances and in the halls can monitor mask compliance and temperature and send alerts to health and safety staff who can meet people who represent a risk and take necessary measures. This is in addition to contactless check-in kiosks and automated boarding systems powered by facial recognition. Buses can also be equipped with health check and access systems using facial recognition, letting-in compliant ticket holders, while denying access to those posing a risk.
Law enforcement has been a key use case for facial recognition since the initial days of the technology’s development. While concerns have been expressed around inappropriate surveillance, bias or false identification, the technology is helpful in certain public sector scenarios. Police has used facial recognition ethically to solve crimes, from identifying shoplifters to tracking dangerous criminals. Public opinion and some mishaps have slowed down adoption of the technology for law enforcement. There is still much work that should be done for ethical implementation of facial recognition, as well as its regulation and controlled use, in order to gain the public’s confidence.
The use of facial recognition by law enforcement and public safety isn’t just about tracking criminals. One positive use case that has seen success where piloted is the recognition of people in need of help but who can’t identify themselves. For example, senior citizens with dementia who are lost and do not carry an ID, or someone passed out in public due to an incident. Law enforcement officers using smartphones equipped with facial recognition can use the technology to match the person’s face with government photo ID records, and provide swift and adequate assistance.
A positive corollary to the above example is the use of facial recognition through cameras located in public places such as busy streets, airports, malls and train stations, to find missing people and victims of human trafficking. As long as their faces are in a database, law enforcement can be notified as soon as they are identified on a camera. In 4 days alone, nearly 3,000 missing children were discovered in India, using facial recognition.
There are many more use cases for facial recognition outside the above categories. Here is a handful of examples worth bringing up:
FaceMe® has been implemented in service robots shaped as humanoids, using facial recognition to enhance and personalize interactions with humans. They can serve as virtual concierges guiding visitors in malls and public libraries. Or they can be programmed to watch over individuals who require monitoring at home for health reasons.
When installed at workplace or school entrances, for example, cameras connected to a monitoring system integrating facial recognition can automatically and precisely track when people come in and out. It can replace time-clock systems at work or keep attendance records at school, all with no human action required.
There have been more instances than the public would care to know, of unauthorized individuals getting on public buses and driving the vehicle around, with or without passengers on board. We have a customer who operates public bus networks and has implemented cameras powered by FaceMe® to verify that only authorized drivers can operate the vehicles. The solution identifies the individual sitting on the driver seat with the database, and ensures it is the right driver, scheduled to operate the bus at that time.
Face recognition can be used to detect certain diseases. The National Human Genome Institute Research Institute is using this technology to detect DiGeorge syndrome, a rare disease in which facial appearance is altered, due to a portion of the 22nd chromosome missing. Facial recognition has helped diagnose the disease in 96% of cases. We can expect many more examples of facial recognition used as a diagnostic tool, as algorithms keep getting more sophisticated.
Facial recognition can play specific roles in casinos, on top of traditional security monitoring. Many jurisdictions keep voluntary exclusion lists to protect compulsory gamblers. Casinos can be imposed hefty fines if someone on the list is caught gambling. Facial recognition can identify list members and keep them away. It can also send alerts to security whenever a known cheater or advantage gambler enters a casino.
Now that we have widely discussed facial recognition use cases, we will provide an overview of its implementation and factors for consideration.
Facial recognition is essentially delivered as a software technology. FaceMe® is a software development kit (SDK). It can be customized and integrated into multiple devices on a network, with key use cases for edge-based IoT and AIoT (Artificial Intelligence of Things.) As an SDK, FaceMe® can be integrated into virtually any device and operating system. Please refer to Facial Recognition at the Edge – The Ultimate Guide 2021 for details on choosing facial recognition software.
On hardware considerations, there is no one-size fits all approach to facial recognition. Solutions should always be customized to the end-user and their industry, as well as the specific use case’s needs and scope.
There are endless hardware options capable of running facial recognition. Some of the most common include PCs, workstations and smart IoT or AIoT devices. For each type of device, there are several options for chipset or camera type which can be used. This is on top of a number of other considerations, each having an impact on performance and the total cost of ownership. For details, please read Facial Recognition at the Edge – The Ultimate Guide 2021.
When designing a facial recognition solution tailored to your specific needs, it is wise to first study well-accepted and widely deployed use cases. For example, if you want to implement facial recognition for employee and visitor access control, you can learn from existing deployments in businesses comparable in size and scope.
It becomes trickier when the use case is new. This is when you want to determine and analyze all the factors and variables first. Some of the most important considerations include:
Once these items are determined, you should be able to better grasp the size and scope of the facial recognition solution that is right for you.
We illustrated the use cases mentioned earlier with examples from a number of industries. When taking a vertical look on the market, there are 10 industries that stand out as being ripe to integrate facial recognition and, in many cases, are already embracing it.
As discussed throughout this article, the use cases for facial recognition are boundless. What’s most exciting is all the future use cases that have yet to be discovered, or are currently being tested and yet to be widely deployed.
As long as the chipset and hardware makers continue to bring better, faster, cheaper products to market, which they will, new use cases for facial recognition will come in the market. A few notable, emerging use cases we are excited about include:
We have only scratched the surface for what facial recognition can deliver to improve safety, security and efficacy of our world. It’s the future of AI and biometric technology. And we are looking forward to continue innovating and delivering our world-class FaceMe® solutions to end-users.
For a full overview of facial recognition, how it works and how it can be deployed, read Edge-based Facial Recognition - The Ultimate Guide.