Face recognition technology is becoming more and more popular in personal devices such as smartphones and laptops. This is unbelievable at first glance. Just a few years ago, most of us used a PIN, password, or even a template to unlock our phone, and now with just a glance, the device is ready to take us to the necessary tasks without a moment’s thought.
However, even with this level of advancement and confidence, you should know that not all face recognition technologies used in different devices are the same. Some methods are more secure than others, and some, by providing safety options, allow us to contribute to the security of this technology.
With that in mind, let’s explore some of the methods of face recognition technology and their differences later in this article. Later in this article, we will discuss security and whether or not to use them. Join us.
Face recognition based on normal cameras
As the name suggests, this technology is based on the front camera of the device and recognizes your face through the lens of the selfie camera. All Android phones have been equipped with this feature since the release of version 4 of this operating system called Ice Cream Sandwich.
So you may not believe it, but this feature existed even before fingerprint sensors and became the first biometric method of unlocking smart devices.
How this works is very simple: When you enable this feature for the first time, the device captures your face image from several different angles. It then uses the software algorithm to extract your facial features and save them for future reference. From now on, as soon as you try to unlock the device using this method, the live image of your face will be compared to the sample that was previously saved.
The accuracy of this method depends entirely on the software algorithm used, so it can be guessed that it is far from perfect. When it comes to variables such as lighting conditions, facial changes (such as makeup), and the use of accessories such as sunglasses, jewelry, and even masks, it can be seen how difficult the device is incorrectly recognizing the face. Also, software-based face recognition technology must strike the right balance between accuracy and speed.
While Android has been providing the APIs needed to use this method for many years, other Android phone companies have been designing and developing their own methods over the years.
It does not matter if Android APIs are used or such methods, the main goal is to increase the speed of face recognition with the utmost accuracy or security. Unfortunately, even with all these efforts, some methods based on software algorithms are deceived by the image of a person and thus question their acceptance.
Infrared based face recognition
Although most smartphones have been able to use software face recognition technology to unlock the device for years, the infrared-based method requires distinct secondary hardware. Of course, here we have to remind once again that even infrared-based methods do not have the same security and efficiency.
The first type of face recognition hardware used a two-dimensional image of the face for its database. This method is very similar to the software method, but with the difference that it uses infrared (InfraRed) spectrum. The main advantage of this method is the elimination of one of the biggest limitations of ordinary cameras in capturing images.
This means that the infrared-based face recognition method does not require your face to be very bright, and you can even use it in dimly lit environments. On the other hand, this method is largely resistant to fraud and deception such as image use because infrared cameras use thermal energy to analyze a subject.
Today, 2D IR or infrared face recognition technology is found in almost all high-end laptops that use Windows Hello technology to perform this process.
For example, Microsoft Surface devices support this technology. You can also use this face recognition technology method to increase security by providing some external webcams such as Logitech Brio 4K which has infrared cameras.
But it is too early to conclude. It is easy to accept that infrared camera-based face recognition technology performs better than regular cameras, but there is an even better way. Apple’s Face ID, for example, has taken the technology to the next level thanks to the use of a set of sensors to capture 3D images.
In this method, thousands of invisible spots are considered on your face to record the maximum possible details of your face. The infrared sensor then measures the depth of your face like a detailed map by measuring and extracting the dots.
The use of 3D systems for face recognition has two major advantages: This method can be used in the dark and it is very difficult to deceive them. Also, while 2D systems based on infrared focus only on heat, 3D systems also consider deep detail.
Clearly, the second method, like the one Apple uses, cannot be avoided without the use of precise 3D modeling in the real world and heat injection.
Comparison of infrared face recognition with face recognition based on normal cameras; which one is better?
Undoubtedly, infrared-based face recognition has much higher security than other methods. Do not forget that even if you are interested in using the basic method, some companies will not allow you to use it in certain circumstances.
In Android, for example, Google Mobile Services confirmation text indicates that this method has the lowest priority of biometric authentication methods. So although you can use this method to unlock your device, banking applications such as Google’s own company do not allow you to use this method. Infrared-based methods, on the other hand, usually have a higher level of safety.
Apple, for example, puts face recognition on par with fingerprints and passwords. As a result, in addition to unlocking the device, you can also use this method to perform financial transactions. Peach Windows, mentioned earlier, has the same policy and allows you to perform the most sensitive transactions in the browser after face-based infrared.
Privacy considerations in face recognition
You are right, your considerations are quite logical. In today’s world, how important data is stored, especially data related to the form that can be used in other situations, is very important and access to them is a matter of much discussion.
Fortunately, you do not need to worry too much about this. Almost all operating systems that support biometric retrieval methods have mechanisms in place to securely store sensitive data such as fingerprint and face information. In smartphones, this information is usually stored in a special part of the chip after encryption.
Qualcomm, one of the largest manufacturers of Android smartphone chips, has set aside a separate section for this purpose. Apple also refers to this part of its chips as Secure Enclave, and has been pursuing such a policy for a long time.
In other words, third-party applications are not allowed to access your biometric data, and none of the attackers are able to extract it from the device in most cases.
Do you have to use face recognition on your phone?
Face recognition is the fastest and easiest way to unlock your device, and especially if the only biometric way is to unlock the device, you will understand this more and more. No matter how easy it is for you to use, it is harder to forge than traditional passwords.
Someone else may easily look at the screen of your device to find out the password or pattern, but this is not the case with the face.
But if you have read the article, you know that algorithm-based face recognition methods and ordinary cameras are not very safe. Smartphone manufacturers use different methods to implement this technology. If your phone manufacturer has set high standards for this method, you can use it with more comfort and ease.
Although infrared cameras have provided a very high level of security in this technology, they seem to be rarer than ever in recent years. Aside from recent iPhone and iPad Pro handsets, many smartphones do not use these separate sensors.
A few years ago, several Android devices, such as the Galaxy S8 and S9, the Pixel 4, and the Huawei Mate 20 Pro, used 3D face reopening, which is no longer the case.
The question may be asked why? In fact, companies’ competition to reduce margins and use smaller apertures in new handsets usually does not allow companies to use too many sensors on the front of their devices.
Therefore, the use of software methods is one of the only methods available in these situations, and they have inevitably moved in this direction. Believe it or not, infrared sensors have simply disappeared during this competition and very few companies pay attention to them.
However, one day these sensors may return to the world of Android. Recently, in addition to fingerprint sensors, many companies have tried to move the selfie camera under the screen. The Galaxy Z Fold 3, as the first Samsung phone with this feature, could indicate a bright future in this field.
Apparently now is the only time we have to prepare for the full return of infrared-based face recognition hardware sensors. But what is the task until then? Finally, should we use face recognition method in our phones or not?
If your device supports the infrared method and has hardware such as Facebook ID, using it is a very prudent action that at the same time keeps you safe. But if this is not the case then maybe your fingerprint sensor is a good idea and if your device does have one, we suggest you use it.
Frequently Asked Questions
What is face recognition technology?
- In this technology, many software and hardware methods are used to recognize the face of the device owner and finally reopen it.
What is the safest way to recognize faces on electronic devices?
- Infrared-based face recognition is the safest way to unlock the device through this technology, which requires a distinctive hardware.
Why does Apple offer the safest face recognition technology today?
- Apple’s face recognition technology, unlike other existing methods, scans the face in 3D, so it is very difficult to deceive the device under these conditions.
Source: Android Authority
