16 Nov Researchers Unlock Smartphones Using Fake Fingerprints
Researchers from the universities of New York and Michigan said that they were able to train a machine learning algorithm to create fake fingerprints that can match a large number of real ones in databases.
To create the fingerprints, known as DeepMasterPrints, the researchers fed an artificial neural network around 6000 real fingerprints from various individuals which was passed onto a “generator” neural net for analysis. From this, synthetic prints are produced and then fed into a “discriminator” neural net which determines authenticity. The process is then repeated thousands of times until the generator is able to fool the discriminator.
The fake fingerprints are said to be specifically designed smartphone fingerprint sensors as these capacitive fingerprints scanners usually only take partial readings of fingerprints when they are placed on the sensor.