Speech sounds and voices can be identified using a unique “neural fingerprint” from an fMRI scan of the listener’s brain, Maastricht University researchers have found.

This new knowledge could be used in the future to improve computer systems for automatic speech and speaker recognition.