What comes to mind when it comes to biometrics? Fingerprint identification has long been popularized; Or the popular face recognition; Or is it more “tall” pupil recognition? Although the biometric information recognized by the current biometric technology is very different, the biggest promise of biometric technology is to replace the past typed security system with new body based identity recognition. Biometric systems no longer need external tokens such as stored passwords or physical key cards, but identify people through their internal characteristics (such as appearance, voice or way people walk). With the continuous improvement of technology, in addition to the functions of face recognition and fingerprint scanning, there is also a person’s unique “heart pattern”, a person’s smell, the subtle skin vibration of his face, throat or chest when talking, and even the shape of his hips. All these have become proof of concept biometric security systems that can confirm the identity of users.
Escalating biometric technology will always surprise people
According to media reports, researchers at the State University of New York at Buffalo have come up with a new method. Although it sounds strange on the surface, it is actually surprisingly practical. Their new biometric tool, called earecho, uses sound waves to identify users according to the unique geometry of the ear canal. This information is very difficult for a person who wants to deceive others.
The team purchased a pair of off the shelf earplugs and made some modifications. A microphone facing the wearer’s ear canal is added to the earplug. That is to unlock the mobile phone by using wireless earplug sound waves, which use sound waves to identify the unique geometry of the ear canal and judge the action you want to make. This paper was published in the Journal of the American Computer Society (ACM) on September 19. It is worth mentioning that the journal is published quarterly by the computer association. At present, the existing data show that about 95% of the research results of the journal may be used in real life.
How does the new technology earecho work
Simplified acoustic modeling diagram of sound propagation and reflection in the ear canal. P is the direct propagation of sound wave, PI is the incident sound wave, and PR is the reflected sound wave
The R & D team used off the shelf products to make prototypes, including a pair of earphones and a small microphone, to make an earecho sample product. Then, the team developed acoustic signal processing technology to limit noise interference, and established a model for sharing information between earecho components.
When the sound is transmitted to the human ear, the sound will be transmitted through the ear canal and reflected and absorbed by the ear canal. All sounds will produce a unique sound transmission that can be recorded by the microphone. The information collected by the microphone is sent to the smartphone through the Bluetooth connection of the earplug for analysis. After acoustic signal preprocessing, the features based on transfer function are fed into SVM classifier for authentication.
Facts have proved that the prototype of earecho device also has amazing sensitivity and can identify users with an accuracy of up to 95%. This was shown in the test, in which 20 subjects listened to audio samples from spoken language to music. It was tested in various locations, such as shopping malls or streets, and subjects stood or sat in different positions. The demo device can identify users with 95% accuracy in one second, and if it can continuously monitor them for three seconds, the score rises to 97.5%.
What are the unique advantages of earecho technology
In some scenarios, such a solution may be useful. The most obvious is the seamless smartphone unlock. Although Apple’s face ID and other systems work well, imagine if your smartphone is on, you just need to wear headphones to know who you are. Isn’t it great? Especially for enthusiasts of Bluetooth headset, it’s perfect. In addition, it can be revalidated continuously. This means that you don’t have to unlock your phone every time you take it out of your pocket.
Another possibility is remote phone based authentication. If you want to make a private call with someone, make sure it’s that person. In the deepfake audio world, this may become more important because it is not impossible to simulate a person’s voice in real time. Using technologies such as earecho, you can easily verify the identity of the person you are talking to. By handing over your in ear information to the bank, you can prevent fraud and stop you from taking the time to answer personal questions that prove who you are.
Compared with the most popular mobile and wearable authentication solutions at this stage (i.e. fingerprint, face ID, voiceprint, etc.), earecho has the following advantages:
The first is to pick up and use in real time. Microphone and earphone speaker are two basic components of wireless earphone design. They are small in shape, low in cost and can be easily taken away. The second is user friendliness. Earecho does not require active authentication. It can automatically process authentication requests when users wear headphones.
The third advantage is that it is not attractive. One of the main problems of fingerprint and face based mobile authentication is to cheat the system after the user information is illegally obtained. Therefore, in the earecho sample, the data collection and authentication process are carried out in a more unobtrusive way. It is difficult for malicious attackers to steal the ear (including physical and physiological) information of registered users through bypass attacks. This information will not be retained continuously, and another information identification will be updated with the next authentication.
The last advantage is reliability, stability and robustness over time. This new technology can deal with the uncertainty of noise, body motion and sound pressure level in different backgrounds.
Although there are still loopholes in biometric technologies such as face recognition and fingerprint recognition, few biometric technologies have not been measured and evaluated in some way in our daily life. Knowing when to use these tools and doubling or even tripling the security through the use of several tools will make our future interaction with machines simpler and safer.
This means that units such as banks that need to accurately identify people will one day ask you whether you want to submit information about the shape of the inner ear.