Gyrophone: Recognizing Speech From Gyroscope Signals
Authors: Y. Michalevsky, G. Nakibly, and D. Boneh
Abstract:
We show that the MEMS gyroscopes found on modern smart phones are
sufficiently sensitive to measure acoustic signals in the vicinity of
the phone. The resulting signals contain only very low-frequency
information (<200Hz). Nevertheless we show, using signal processing
and machine learning, that this information is sufficient to identify
speaker information and even parse speech. Since iOS and Android
require no special permissions to access the gyro, our results show
that apps and active web content that cannot access the microphone
can nevertheless eavesdrop on speech in the vicinity of the phone.
Reference:
In proceedings of Usenix Security 2014
Full paper: pdf