6533b82ffe1ef96bd1294fcc
RESEARCH PRODUCT
On Information Processing in the Cat’s Visual Cortex
Hanspeter A. MallotW. Von SeelenH. R. O. DinseG. Kronesubject
Basis (linear algebra)business.industryOrientation (computer vision)Computer sciencemedia_common.quotation_subjectInformation processingPattern recognitionTask (project management)ControllabilityObservabilityArtificial intelligencebusinessFunction (engineering)Complement (set theory)media_commondescription
We assume that the visual system serves for orientation in space, recognition of objects and the interpretation of scenes and scene sequences. This task breaks up into a series of partially interdependent subproblems which are solved by some 13–15 usually retinotopically organized areas. So far it has not been possible to correlate functions and areas unequivocally. One reason for this could be the inadequacy of the questions posed as a basis for experiments. However, we think it more likely that correlating a function with an area is, as a rule, inadmissible since the degree of the coupling in the whole system does not permit a simple divsion. Rather the type and degree of coupling determine the separation or integration of “elementary units”. It follows that co-operation between subsystems is an essential aspect. The data available to us from neurophysiological, neuroanatomic and behavioral physiological findings is insufficient to analyse such a system: they ensure neither the complete observability nor the complete controllability of the systems. So one is forced to use model comparison for analysis. Depending on the type of data, we use three models which complement each other to order the data and predict results which can be experimentally verified.
year | journal | country | edition | language |
---|---|---|---|---|
1986-01-01 |