Today I want to talk about one of the “hot topics” within the security camera and surveillance industry- facial recognition.
As security cameras become smarter one of the “Holy Grails” has been to develop a camera that can recognise people by their faces.
The smarter that cameras become, the less work humans have to do.
And if a camera can identify individuals, how awesome would that be?
It would save hours of effort trawling through footage as instead of being warned every time a person is detected, an operator is only warned when a stranger or known offender is filmed.
But what about our privacy? Our right to not be be constantly monitored every minute of the day? And just how good is the technology anyway?
What is Facial Recognition?
Facial recognition technology enables a digital device to recognize a human face, analyze it, and compare to an existing database of known faces. When it finds a match, it can help verify a person’s identity.
With its high-tech features, facial recognition has several commercial applications. The technology is so versatile that it can be used for monitoring, surveillance, security, access, and even marketing. And because businesses are finding more ways how to use facial recognition in their operations, people can expect more applications of the technology.
How Facial Recognition Works
Facial recognition sounds like a next-generation technology with futuristic applications, but the truth is that it has been around for more than five decades. But since its invention in the 1960s, it has become a man-machine of some sort because of high-level artificial intelligence that can mimic processes that only the human mind can do.
How the technology works depends on where it is used for, but at the basic level, facial recognition works in simple steps:
- A face is detected through a video or photo capture. The face may appear as part of a large crowd or as a solitary image. The device does not see it as a “face” in its strictest sense, but a bunch of data that represents the set of features of the face. Think of it as the fingerprint of the face—a faceprint, if you will. This data is essentially the geometry of the face.
- The facial recognition software reads the data and tries to make sense of it by analyzing the face features with respect to head rotation, tilt, angles, scale, position, and other metrics for identification. The algorithm is able to identify features that distinguish the face from the other faces the device captures. When it processes the data, the resulting image is a unique facial signature.
- The software then takes the facial signature and compares it to a database. The type of database would depend on what institution is using facial recognition technology. For instance, the FBI has a database of known faces which has over 640 million photos of American citizens.
- The facial recognition software determines whether or not the face in question matches an image in the database. It will alert the user if a match is found or not.
That’s just the simplified version of what is actually a complex process using sophisticated algorithms. Companies are now using deep learning to interpret the huge amount of data from videos and images that come from devices like smartphones and security cameras.
Users of Facial Recognition
Facial recognition is used at varying degrees by different organizations. It’s not limited to surveillance, crime prevention, and public space monitoring. You’d be surprised how the technology is being utilized by different sectors of society to make life a safer and more convenient.
- Mobile phone manufacturers – Smartphone makers use face recognition to authenticate users of the phone. Aside from password, pattern, and fingerprint scan, face ID can be used to unlock phones and tablets. It is said to be an effective method because the chance of the phone getting unlocked by a random face is one in one million.
- Airports and airlines – In partnership with the US government, airports are using facial recognition technology to monitor people that are coming in and going out of the country. It makes it easier for the Department of Homeland Security to keep an eye on people who have expired visas or those under criminal investigation. In fact, with the use of biometric technology at the Washington Dulles International Airport, a man with a fake passport was arrested by customs officers. It was the first arrest using facial recognition. Airlines have also installed special cameras at departure gates to scan faces of passengers.
- Companies – Traditional IDs or security badges will soon be a thing of the past when facial recognition systems become more widespread. Many tech companies are utilizing fingerprints faceprints as authentication methods to gain access to company premises and restricted areas.
- Religious institutions and places of worship – Church groups are getting digital nowadays. Megachurches are using facial recognition to monitor their congregations’ attendance to improve donation collection. It is also used for security purposes considering houses of worship are becoming targets of terrorist attacks.
- Retail stores and commercial establishment – Retailers are losing big money from shoplifting and employee theft. This prompted them to improve their monitoring and surveillance effort by adding face recognition to their security arsenal.
- Schools, colleges, and universities – Schools are making tremendous efforts to make campuses safe for students and teachers. Face recognition enables the identification of unauthorized visitors and alerts the security staff and school personnel.
- Social media platforms and apps – Some apps and are able to detect faces across multiple social media platforms. Perhaps the most common application is Facebook. It detects faces from uploaded photos and notifies you that it someone from your network. It would then ask you if want to tag that person. What’s amazing is that Facebook’s facial recognition technology is accurate 98% of the time.
- Marketers and advertisers – Facial recognition advertising is a thing and companies are using it to target groups to promote their products, services, and ideas. Face detection cameras can pick out the gender, age, emotional response, and other data. These data are then used to tailor advertisements to match the interests of individuals and groups.
These are just some of the applications of facial recognition. There are more in the pipeline and tech companies are working to improve the technology.
Security Cameras with Facial Recognition
Clearly, facial recognition is a powerful tool used to tighten security in critical places. Although it is past the infancy stage, the technology is not as widely used at the consumer level.
Only a few security camera brands have facial recognition features and they are not as sophisticated as those used in public spaces. Advanced features tend to drive up the price so not all manufacturers are keen on adding the facial recognition feature in their entry-level and mid-range models.
Of course, there’s the influx of cheap security cameras from China, but the quality is quite dodgy, to say the least. Oftentimes, the facial recognition software is a hit or miss.
Perhaps the consumers themselves don’t see the need for such an advanced feature in their home security cameras. But that can change in the near future because big tech companies are now integrating the technology in smartphones, mobile devices, and home systems.
Here are some notable face recognition security cameras and home systems that are making headway in the consumer market realm.
- Nest Hello
Nest Hello is Google’s answer to Amazon’s Ring video doorbell. There’s no runaway winner when it comes to performance because each has its own strengths. However, Nest Hello has the upper-hand in the face tracking domain.
With its Face Alert feature, Nest Hello is able to detect and recognize faces and lets you know who those faces are from the database of known faces. If it’s unable to identify the face, it’s likely that there is a stranger, an unwanted guest, or a burglar.
The feature is only enabled if you subscribe to the Nest Aware cloud service, which is a monthly subscription plan. Nest Hello is cheaper than Nest Cam IQ (Indoor and Outdoor) but it has a more impressive facial recognition feature because it gives more information when it comes to identifying who’s at your front door.
- Tend Secure Lynx
Lesser known Tend Secure Lynx is flexing its facial recognition muscles and getting a lot of attention largely due to its affordable pricing. Hovering in the $60 price point, it is one of the cheapest in its security camera class.
Once you have created your database of known and authorized faces, the camera will familiarize itself with the faces and prompt you when it matches the face at the door or within the security camera’s viewing angle.
- Netatmo Welcome
Netatmo is one of the early adopters of face recognition technology and it implements the feature to mildly satisfactory reviews. The camera’s built-in tracking system does not require cloud subscription for the feature to work, so that’s a big plus for the budget-conscious.
Netatmo doesn’t quite compete with the best of the lot. It’s less responsive and has occasional lag issues, but it does a fairly decent job of detecting and recognizing faces and enables customization. Next-generation models will most likely address the issues.
- Wisenet SmartCam N1
With a $50 price point, this wireless indoor security camera is clearly a budget model. But amazingly, it performs pretty well in identifying faces. It works seamlessly with motion detection features and has support for Google Assistant and Alexa.
When it detects human motions or identifies a face from the database, it sends alert notifications. It’s by no means perfect, but the ability to identify faces improves over time with continued use.
- Google Nest Hub Max
While this is not a security camera in the strictest sense, it functions as one. It may be considered a technology that is always watching everyone and everything around. Google’s foray into the facial recognition territory is creating an avalanche of tech devices with the same Big Brother features.
Nest Hub Max is a Home Hub on steroids. It not only works as a speaker, screen, and smart home controller, but it also functions as a facial recognition device. With the added camera on the device, it becomes an all-seeing eye of sorts. It’s a smarter Google Assistant with the ability to detect, recognize, and identify a person inside your home.
When the camera sensing feature is enabled, the device stays on and looks for familiar faces. When it recognizes your face, the algorithm will go to work and the device will start to interact with you by showing you a quick rundown of your day.
When another person is spotted, the device interacts with different information based on what it knows about the person. It acts like an intelligent robot that can understand context and it could tailor-fit information to match your interests. It’s amazing but it’s also creepy and unsettling.
The Hub Max camera works as a Nest Cam. It can send you an alert when it spots a familiar face or a stranger.
Google assures that all these advanced features are “opt-ins”, meaning that you can choose not to use the face match features. Google makes certain that the face-sensing processes only happen locally on the device itself.
Images and videos are deleted immediately and are not stored anywhere. Furthermore, Google does not keep a record of what the camera detected. Data cannot be used for marketing. These should allay some privacy fears from using the technology…for now.
It appears that companies are taking a hands-off approach when it comes to facial recognition and the likely reason is security breach and privacy issues. As it stands, there is a lack of federal regulations surrounding the use of the technology. Without clear-cut rules on how to responsibly use facial recognition, it will remain an iffy subject.
Another cause for concern is the technology’s accuracy (or inaccuracy) in identifying people with similar facial features. It could lead to misidentifying people as suspects to crime and can result in wrongful convictions.
Apart from technology issues, there’s the touchy subject of community acceptance, government legislation, and security concerns. These are critical issues that must be sorted out because using the technology entails collecting biometric data from people, with or without their knowledge.
The best practice is to always inform people if their biometrics are being collected and being matched to a database. It’s just the ethical thing to do, but without laws in place, who’s to say that companies are being ethical in their business practices?
States and cities that limit the use of facial recognition in law enforcement acknowledge that efforts to keep communities safe should not be at the expense of fundamental rights and freedoms of people.
The Future of Facial Recognition
Companies are getting more creative in how they plan on using facial recognition technology. This is why the global market for facial recognition is expected to grow from $4.05 billion in 2017 to $7.76 billion in 2020.
The growth will likely be spurred by the growing need to enhance monitoring and surveillance at public places, largely due to smart city initiatives of governments. Not far behind are companies that are taking the technology to mass consumer markets.
Facial recognition technology is not yet perfected. There is a tremendous room for improvement and tech companies are exploring how the technology can be put into good use without causing fear and panic.
Worst-case scenario: Face recognition technology will be abused by people who would use the data for criminal and perverse activities. It’s not hard to picture what sinister plans hackers and criminals have on the pipeline.
Best-case scenario: Face recognition technology will be perfected to provide accurate identification and legislation will be put in place to protect privacy rights.