3D facial recognition is collectively abandoned by Android manufacturers, will under-screen fingerprint take over?

The 3D face recognition technology that was popular last year has disappeared from the new Android phones this spring. With the continuous pursuit of larger screens and higher screen-to-body ratios, under-screen fingerprint recognition technology has become the "mainstream" biometric recognition option.

New spring phones all abandon 3D face recognition

Around the 2019 Mobile World Congress, many major brands chose to release new Android flagship phones during the period, and these new products at least represent the trend of Android phones in the first half of 2019. However, judging from the Xiaomi 9 and Samsung S10 series that have been released, as well as the Samsung Galaxy Fold and Huawei Mate X with folding screens, perhaps under the competitive pressure of screen-to-body ratio, these new Android flagships have not chosen to support 3D face recognition. In addition, it is worth mentioning that many upcoming mobile phone products, judging from their appearance alone, generally adopt a water drop screen design, which indirectly indicates that the probability of them supporting 3D face recognition is very small.

However, in the same period last year, Android phone manufacturers still took an active follow-up attitude towards 3D face recognition or its derivative applications. For example, in the Samsung S9 product, although 2D face recognition is still used, a new recognition solution is adopted. By integrating face and iris recognition functions, more accurate recognition can be performed with the help of new hardware, and the AR Emoji application is launched. However, this year, Samsung was limited by the hole-punch screen and abandoned the iris recognition function in the S10 series, which basically means that there is only retreat and no progress on the road of face recognition.

Xiaomi is also similar. Last year, the Xiaomi Mi 8 smartphone launched an Explorer Edition, which achieved 3D face recognition based on coded structured light technology. It also released a set of 3D moji promotional videos similar to the iPhone Animoji style. However, the subsequent progress of this function was also not reflected at the Xiaomi Mi 9 launch conference. Both the Xiaomi Mi 9 Transparent Premium Edition and the Standard Edition only provided ordinary face recognition unlocking function. There has been no news about 3D moji since the second half of last year.

2D face recognition has obvious flaws

If Android phone manufacturers are ready to give up their efforts on 3D face recognition this year in pursuit of a larger screen-to-body ratio, then the face recognition function of Android phones will seem to have returned to the level of a few years ago. Tests have shown that Xiaomi 9 can be unlocked through photos because it does not use advanced face recognition technology. In other words, returning to ordinary face recognition technology in Android phones has greatly reduced the security of face recognition, returning to the level when this technology was first used in mobile phones and could be deceived through photos.

In addition to the fact that 2D face recognition is difficult to guarantee in terms of security, as Apple has begun to gradually cultivate the ecosystem of 3D face recognition in iOS applications, the user experience of Android face recognition applications will lag behind significantly. For example, in the generation of personal 3D portrait emojis, the portraits generated by iPhones are more similar to the user than those generated by Android phones. The reason is that the 3D structured light of the iPhone can capture more facial information, making the image more three-dimensional. In addition, taking the Xiaomi 8 Explorer Edition as an example, although it supports 3D structured light, the face recognition payment function of the phone has not been realized, so the so-called "black technology" only stays at the unlocking stage in application.

3D face recognition has also given some businesses an opportunity to market through apps. Last year, the luxury brand GUCCI added the Animoji feature to the GUCCI app in order to match the company's Boston Terrier-style shoulder bags and clothing. Users can play with two virtual dogs named Bosco and Orso through Animoji in the app and share their expressions with others. At the time, this move also quickly boosted the download volume of the GUCCI app and increased the exposure of supporting products. However, if you want to play with the dogs, you couldn't use it normally if you didn't have an iPhone X that year.

The successor's under-screen fingerprint has some minor problems

However, in terms of biometric security, Android phone manufacturers have obviously found an alternative solution, and under-screen fingerprint recognition has become the central choice for Android phone manufacturers to solve the biometric problem while expanding the screen-to-body ratio. Judging from the new flagship products in spring, unlike last year's gradual exploration of under-screen fingerprint recognition applications, under-screen fingerprint recognition technology has become standard.

Although the recognition speed of under-screen fingerprint recognition is not much different from that of traditional capacitive fingerprint recognition through optimization, it is still criticized by some users for its aesthetics, that is, the fingerprint recognition module of mobile phones with under-screen fingerprint recognition technology can be seen under strong light. In fact, the Samsung S10+ in the picture above was not aesthetically accepted by some users because the recognition module was not placed to the left.

In addition, the use of ultrasonic fingerprint recognition technology also poses a challenge to the later maintenance of the phone. Existing tests have shown that mobile phones with ultrasonic fingerprint recognition technology are picky about the film, and many films will cause the fingerprint to be unrecognizable. However, if the film is not applied, once the screen is severely damaged, the cost of replacing the screen will be too high. ***'s S10+ disassembly report shows that since the ultrasonic under-screen fingerprint recognition solution is fixed to the bottom of the screen with glue, the sensor itself cannot be close if the screen is not removed, and this process will undoubtedly damage the screen. Ultimately, the cost of repairing the phone has increased significantly, which is not good news for users who currently have a general replacement cycle of nearly 24 months.

3D face recognition and under-screen fingerprint are both big cost drivers

In fact, the consumer market is also looking forward to mobile phones that can support both under-screen fingerprint recognition and 3D facial recognition, so that the convenience of mobile phone biometrics can be greatly improved. On the one hand, when it rains and your fingers are wet, 3D facial recognition can solve the problem. At the same time, in foggy weather, fingerprint recognition can also quickly unlock the phone when wearing a mask.

On the other hand, both 3D face recognition and under-screen fingerprint recognition currently occupy an important position in the BOM cost of mobile phones, indirectly affecting the final selling price of mobile phones. In the disassembly cost of Samsung S10+, the 6.4-inch Infinity-O super-sensory full-view screen (including ultrasonic fingerprint component) is worth $86.5, which is the most expensive component in the entire phone. The screen component cost of the previous generation S9+ is $72.5.

The cost of 3D face recognition is not cheap. Last year, the cost issue became a difficult choice for Android phones when choosing whether to adopt 3D face recognition technology. At that time, if you wanted to purchase related 3D sensors to achieve the same effect as iPhone X's Face ID, the cost might be as high as RMB 200 to 300. This will undoubtedly have a significant impact on the pricing of Android phones.

In addition, in order to cooperate with biometric technology, mobile phone SoCs have also begun to add NPUs to improve the intelligence of recognition, such as the neural network function added to the A11 chip of iPhone X at the same time as Face ID. This also means an increase in the cost of mobile phones. Of course, the addition of NPUs also adds more AI features to other functions of mobile phones.

At present, from the perspective of the Android system itself, Google has already added native support for 3D face recognition in Android Q, and will also allow users to use facial recognition to authorize purchases and open applications. Obviously, Google does not want 3D face recognition to become a missing feature in the Android camp. It is expected that Android Q will be pre-installed on newly released mobile phones in the second half of this year. By then, it is unknown whether new Android phones with native support for 3D face recognition in the system will be able to welcome the return of this technology.