CNS Seminar

Although the idea of perception as an inference problem goes back to Helmholtz, it is only recently that we have seen the emergence of neural models of perception that embrace this idea.  Here I shall describe two neural models of perceptual inference that give us new ways of thinking about response properties in primary visual cortex:  1) sparse coding, in which neurons compete to infer a compact description of image data in terms of a set of basis patterns, and 2) dynamic routing, in which separate neural populations representing form and motion interact to infer a stable representation of the world in the face of eye movements that occur during fixation.  In contrast to traditional computational models based on deductive processes such as feature detection and classification, these inferential computations rely heavily upon recurrent computation in which information propagates both within and between levels of representation in a bi-directional manner. The inferential framework shifts us away from thinking of 'receptive fields' and 'tuning' of individual neurons, and instead toward how populations of neurons interact to perform collective computations.