Teacher
Professional
- Messages
- 2,669
- Reaction score
- 819
- Points
- 113
Scientists have learned how to recreate fingerprints by sound using PrintListener.
A group of American and Chinese scientists have developed the PrintListener technique, which is a new way to restore the unique pattern of papillary lines of fingerprints. The method is based on the analysis of sound characteristics that occur when a finger slides across the touch screen.
Fingerprint protection technologies are extremely widespread and highly trusted today. The fingerprint authentication solutions market is expected to reach almost $ 100 billion by 2032. At the same time, there is a growing awareness that intruders may attempt to steal other people's fingerprints.
According to the results of testing, the researchers claim that they can successfully attack "up to 27.9% of partial and 9.3% of complete fingerprints within five attempts with the maximum security settings FAR (False Acceptance Rate) 0.01%". This is, according to them, the first work using sliding sounds to get information about fingerprints.
According to the study, "finger-rubbing sounds can be intercepted by intruders online with a high probability." To do this, a hypothetical attacker just needs to gain access to the microphone of the victim's mobile device while using popular applications such as Discord, Skype, WeChat, FaceTime, and others that use swipes.
However, the developers of PrintListener encountered a number of difficulties that they had to overcome. In particular, the problem of the weak sound of finger rubbing on the screen was solved using spectral analysis methods. A statistical analysis was also performed and a heuristic algorithm was developed to filter out the physiological characteristics of a particular user.
Using the PrintListener sound analysis technology, the researchers recovered synthetic PatternMasterPrint fingerprints, and "in realistic scenarios" this technique was able to successfully restore partial fingerprints in more than one out of four cases, and full ones in almost one out of ten. The results were higher than the MasterPrint attack, which offers a random search of "universal" prints.
A group of American and Chinese scientists have developed the PrintListener technique, which is a new way to restore the unique pattern of papillary lines of fingerprints. The method is based on the analysis of sound characteristics that occur when a finger slides across the touch screen.
Fingerprint protection technologies are extremely widespread and highly trusted today. The fingerprint authentication solutions market is expected to reach almost $ 100 billion by 2032. At the same time, there is a growing awareness that intruders may attempt to steal other people's fingerprints.
According to the results of testing, the researchers claim that they can successfully attack "up to 27.9% of partial and 9.3% of complete fingerprints within five attempts with the maximum security settings FAR (False Acceptance Rate) 0.01%". This is, according to them, the first work using sliding sounds to get information about fingerprints.
According to the study, "finger-rubbing sounds can be intercepted by intruders online with a high probability." To do this, a hypothetical attacker just needs to gain access to the microphone of the victim's mobile device while using popular applications such as Discord, Skype, WeChat, FaceTime, and others that use swipes.
However, the developers of PrintListener encountered a number of difficulties that they had to overcome. In particular, the problem of the weak sound of finger rubbing on the screen was solved using spectral analysis methods. A statistical analysis was also performed and a heuristic algorithm was developed to filter out the physiological characteristics of a particular user.
Using the PrintListener sound analysis technology, the researchers recovered synthetic PatternMasterPrint fingerprints, and "in realistic scenarios" this technique was able to successfully restore partial fingerprints in more than one out of four cases, and full ones in almost one out of ten. The results were higher than the MasterPrint attack, which offers a random search of "universal" prints.
