“Spiking network optimized for noise robust word recognition approaches human-level performance and predicts auditory system hierarchy” read more