There is a conflict between "wearing a mask" and facial unlocking.
At present, most smartphones provide face unlocking function. However, due to the current severe epidemic situation, it is necessary to wear a mask if you have to go out. Under such conditions, models that only support face recognition, such as iPhone, are facing a major use problem. The mobile phone can't recognize the user wearing the mask, which makes it impossible to unlock the mobile phone successfully, because the touch ID function has been cancelled since iPhone X, which makes the user have to manually enter the password to unlock it.
In the era of wearing masks, can iPhone still complete face recognition?
The reason why the iPhone with 3D face recognition is difficult to unlock after the user wears a mask is because the security of Face ID recognition is quite high, and the user's facial features can be instantly perceived through 30,000 points, thus ensuring that the recognized face will not be deceived by photos and related imitations. However, some models of 2D face recognition scheme can only recognize 128 face feature points, so a photo can be fooled.
Even if some users wear masks, smart phones can recognize the success of unlocking. Generally speaking, there are three main reasons. First, some users use some Android models, and there is no 3D face recognition technology, so the security is low. In the past, even a photo could be used to unlock them. The second is that some users wear masks in an irregular way when using the iPhone. For example, their noses are exposed, and through the learning ability of neural networks, there is a certain chance to unlock them successfully. Third, some iPhone users' mobile phones can indeed identify the owner wearing a mask, but at present, the number of successful users is low.
However, the method of adding Face ID to the mask in the iPhone settings rumored on the Internet is actually a rumor. When users wear masks to enter Face ID, the system will prompt them to show all their faces, which means that they can't successfully enter with masks.
Can you learn to recognize your mobile phone with a mask?
The reason why some users think that the iPhone may recognize the owner wearing a mask through self-learning ability is because the iPhone has a built-in neural network engine in the mobile phone chip. For example, the neural network engine in the latest A 13 chip has eight cores, which can increase the speed by up to 20%. At the same time, two new machine learning accelerators are added, which can perform matrix mathematical operations at the highest speed of 6 times.
In other words, if an iPhone user suddenly wears glasses one day, the mobile phone may fail to recognize for the first time, and the user needs to enter a password to unlock it, and then the machine learns the changes in the user's face. The face can be unlocked when the user turns on the phone next time.
So there is a saying on the Internet that as long as you unlock the mask with iPhone and enter the password 50 times, the machine can successfully learn the face of the owner wearing the mask, and the user needs to wear the same mask throughout the process. But many people try to input it 50 times or even 100 times and find that the mobile phone still can't recognize the face of the owner wearing the mask. However, some users find that their iPhone can learn to recognize what it looks like after wearing a mask.
According to the analysis, the reason for this situation may be that everyone's facial area is different, which leads to the different scope of the mask. Once the identified area reaches the effective area, it is possible to identify successfully. Previous cases show that if users wear sunglasses, it will also trigger the Face ID machine learning mechanism of iPhone, and the size of sunglasses lenses also determines the number of times of learning. However, after the research is completed, if the user wears sunglasses again, but does not wear a hat, he can basically recognize it. Once he wears a hat to cover his forehead, he is more likely to ask for a password, and he will not continue to learn the face of the owner who wears a hat and sunglasses. This also indirectly explains the phenomenon that some users mentioned above find that exposing their noses can greatly improve the recognition success rate despite wearing masks.
Finally, it is not difficult to find that for security reasons, the mobile phone has a default effective facial recognition area, so that a sufficient number of recognition points can be collected. If you want the iPhone to learn to know the owner who wears a mask, the chance of success is very small.
How to unlock faster?
Therefore, instead of unconsciously entering 50 or even 100 passwords into the iPhone, it is better to save the country by curve to unlock the phone temporarily and faster.
For the sake of safety, the power-on password is still needed, so it may be changed to a simple digital password. In addition, after swiping on the wake-up screen, directly press the word "Face ID" on the screen, and the phone will directly enter the password input interface, which can save a lot of useless identification time.
In two commonly used payment applications, WeChat and Alipay, you can temporarily turn off the "brush face payment" function in the settings, and directly enter the payment password for payment during use. There is no need to wait for two facial recognition failures before entering the payment password.
In addition, if you can't accept the inconvenience of only entering the password to unlock after wearing the mask. If you have old models such as iPhone 8 and iPhone 7 with touch ID at home, you may wish to use them temporarily to facilitate going out. Due to the existence of iCloud, complete interoperability and migration between devices is not a problem.
Let's use Android phones for the time being. After all, most Android phones have basically given up the follow-up of 3D face recognition and adopted the solution of fingerprint recognition under the screen. At the same time, it also has a face unlocking function. Although 2D face recognition is not safe, many users have confirmed that wearing a mask can indeed identify the owner.