Researchers at the University at Buffalo have developed a smartphone application designed to stop voice hacking, a security flaw in many mobile apps that enables attackers to use only a few minutes of audio samples to replay a user's voice and trick digital security systems.
The researchers say a prototype of the system proved highly accurate in stopping machine-based voice impersonation attacks. They say voice recognition could become a more common security tool because more Internet-connected devices are being developed that do not have keypads.
Voice-recognition attacks can synthesize the user's voice, but these are detectable by existing algorithms. However, replaying the user's actual voice has, until now, been an effective way to infiltrate digital security systems. Any replay must be broadcast on a speaker, which produces a magnetic field.
The new system uses the magnetometer in a phone, which is normally used for the phone's compass, to detect a magnetic field.
From University at Buffalo News Center
View Full Article
Abstracts Copyright © 2017 Information Inc., Bethesda, Maryland, USA
No entries found