Cortical computational maps for auditory imaging
Neural Networks
Introduction to Reinforcement Learning
Introduction to Reinforcement Learning
Learning to Predict by the Methods of Temporal Differences
Machine Learning
| Hi-index | 0.00 |
Vocal learning in humans is isolated in evolution, nevertheless the comparative approach can help give insight into mechanisms of speech. Birdsong and speech learning share similar computational problems and may share similar neural mechanisms. In birds, there is a well-known sensory representation of the individual's own song. A similar representation is postulated for speech, and the implications are explored. Auditory responses to own song are observed in an anterior forebrain pathway that is implicated in song recognition. By analogy, speech perception may be referenced to an internal representation of own voice. This also predicts sex differences in speech perception. In awake birds, auditory responses are reduced or absent in motor pathways, which argues against a theory of perception by reference to production. Recent evidence demonstrates that auditory responses in birds are much stronger while animals sleep, and that during sleep there is replay of song motor patterns. If humans exhibit similar phenomena, then there should be circadian patterns in speech production, and possibly in speech perception. Humans should exhibit strong responses (e.g., ERP) to playback of speech during sleep when playback is matched to the perception of own speech. Sleep should be important for speech maintenance, and learning.