Electronic Ears

Sound recognition software can analyse audio in ways that our hearing cannot. But human hearing has the edge in other skills – discriminating one audio source from a background of many for example, like a conversation at a loud party. Which is why engineers at the UK’s Newcastle University are developing a computer model of the ‘auditory midbrain’, the lump of grey matter that deals with the way we perceive sound. The results could enable voice-commanded robotics and a new generation of hearing aids.