Eye pattern recognition is here too. What technologies are currently available to collect our biometrics?

In the 2002 movie Minority Report starring Tom Cruise, criminals used the iris identity left by Tom Cruise to send him a large number of pictures and messages, interfering with his normal work. Such scenes were classified as science fiction plots at the time.

Iris recognition is now a reality, and a few days ago, the white of the eye was also announced to be used to identify people, even identical twins can be distinguished. This eye pattern recognition technology was developed by EyeVerify, a biometric technology startup, which was later acquired by Ant Financial for $70 million and renamed Ant Zorro in October this year.

Eye lines are the arrangement of blood vessels in the white area of ​​the eye. Each person's eye lines are unique, so they can be used as a biometric feature to identify a person. According to reports, under sufficient visible light, users can naturally look at the front camera of the phone to perform eye line recognition, without the need for a special camera like iris recognition. However, this technology has not yet solved interference factors such as eye reflection, blinking, eyelashes, etc., and is still in the laboratory product stage.

In contrast, iris recognition has been used in real life. Samsung Galaxy S8 has built-in iris recognition technology, which requires users to appear in front of the camera in a specific posture for the system to recognize them. This is not convenient enough, resulting in its low practicality.

Iris recognition on the Samsung Galaxy S8.

Fingerprint recognition is the most common technology in our daily life. In 2013, fingerprint recognition first appeared in iPhone 5s, and many people called it a "useless design", but now it has become a security line of defense for most mobile phones. According to data provided by Fingerprints, a total of 1.6 billion mobile phones were sold worldwide in 2016, and 60% of them have fingerprint recognition technology.

Now, the Swedish biometric technology company FPC (Fingerprint Cards) is researching a technology called under-screen fingerprint recognition. The fingerprint recognition module is placed under the mobile phone screen, and the fingerprint pattern is detected and compared through the signal reflected by ultrasound or active infrared light, thereby completing the steps of identifying identity and unlocking the phone.

Mobile phones equipped with this technology can recognize fingerprints anywhere on the entire screen. In addition, it has no rigid requirements on the screen material, and both LCD and OLED screens are acceptable. If this technology is mass-produced, it will be an improvement in convenience and manufacturing cost.

Apple iPhone X was released a few days ago, and Face ID has become popular.

Apple's facial recognition is achieved through a small area on the top of the iPhone X that is not covered. The infrared lens, flood illuminator and dot projector here can project invisible light onto the face to form a 3D face map attached to the user, and compare it with the face map of the owner stored in the phone. If they match, the phone can be unlocked. Apple said that the false match rate of Face ID is one in 1 million, which is 1/20 of Touch ID.

Apple demonstrates how Face ID works.

In fact, if the accuracy requirement is not so high, the cost of facial recognition can also be very low. It is said that the current licensing fee for low-end products with facial algorithm deployment in the market is about 500 yuan per set. Some time ago, Gome launched the S1 smartphone with facial recognition, which was priced at only 1,299 yuan; ZTE launched the Blade A3, which was priced at 799 yuan.

Human voice is also used as a biometric identifier because of its unique characteristics. Its most common application is personal privacy protection. For example, WeChat has a built-in "voice lock" function to protect user accounts from being stolen.

But its greater use is in smart speakers. Imagine if your smart speaker can accurately recognize the voices of more than a dozen people at home, it can play your grandfather's favorite song "Red Sunset" when he speaks to it, and turn on the light in your mother's room when she goes to bed.

Voiceprint recognition is still in its early stages. Amazon's Echo launched the Echo 2nd generation in October this year, which can recognize more than a dozen voices. Domestic Tmall Genie and Ding Dong speakers also have voiceprint recognition functions.

Palms are also used as identity verification tools. Some time ago, the US startup Redrock Biometrics announced that they have developed a technology that can complete identity verification by "waving your hand". It uses the uniqueness of a person's palm prints and can complete the verification with an ordinary camera and CPU processor. The whole process only takes between 10 and 100 milliseconds.

However, these biometrics face privacy issues due to their low accuracy.

After the Samsung Galaxy S8 came out, the German hacker website Chaos Computer Clubs conducted an experiment. They used a 200 mm lens to take a photo of a person's iris at a distance of 5 meters, fine-tuned the light and dark contrast of the photo, and then used a laser printer to print out the person's eye pattern, and then pasted contact lenses to imitate the curvature of the eyeball. The experimenter used the fake glasses to face the iris recognition lens of the Samsung Galaxy S8, and the phone was unlocked.

New York University and Michigan State University also released a report that found that smartphones can be easily fooled by electronic fake fingerprints. In the experiment, university researchers developed a series of simulated fingerprints called MasterPrints, which had a match rate of 65% with real fingerprints that can be recognized by mobile phones.

Sample fingerprints used by researchers at New York University and Michigan State University to compare with the so-called MasterPrint. The fingerprints with a solid black line around them are the ones that have been successfully matched with the MasterPrint.

People are also worried that mobile phone companies will take away their personal information. Three days after the release of iPhone X, Apple CEO Cook received a letter from US Senator Franken, Chairman of the Privacy, Technology and Law Judiciary Committee, questioning the protection of facial data. In response, Apple said that the entire recognition process is carried out on the phone and the company does not own the user's facial data information.

All the parts of the body with unique characteristics are used for identification. In the final analysis, it is because mobile phones store more and more personal information, as if they have become a part of people. Excessive intimacy will bring privacy issues. This seems to be a vicious circle, but people are still trying to solve it.