Linking Man and Machine Through Adaptive Sensory Processing.

摘要

Our project aimed to understand the strategies used by sensory cortex to adapt to complex environments, and to signal the appearance of novel events in those environments, This knowledge can be used to develop novel computational algorithms that employ the same strategies. The algorithms can be used to better integrate man and machine. In the first year of our grant we have made substantial progress in understanding some of the biological markers of novelty detection in both human subjects (EEG) and in single unit recordings in nonhuman primates. In the second year of our grant, we finished data collection and developed a computational algorithm that mimics this type of learning or adaptation. Together, our results and modeling reveal new strategies by which the visual and auditory systems adjust to statistical regularities in complex environments, allowing them to signal the occurrence of deviant or unexpected events.