Inspired by the fact that even the simplest of animals, such as insects, have remarkable perception capabilities which are difficult for present-day engineered systems to duplicate, this project seeks to understand how the brain accurately organizes and processes on the macro-scale level the temporal flow of visual information at a lower stage of the visual pathway that is the primary visual cortex (also called area V1) and to develop machine vision systems based on the findings to mimic brain visual function. The novel aspects of this project include the innovative use of scalp electrodes to probe macro-scale processes of the visual cortex, investigation of encoding of sequences of patterns rather than single presentations, use of innovative mathematical models, and recognition of the existence of several different, task modulated, low-level perceptual mechanisms. The experimental part of the project focuses on the storage of temporal patterns, presented continuously over the time scale of tens of minutes, and how the stored information is integrated in the primary visual cortex with the newly acquired visual input. Scalp recordings will be used to investigate collective responses of neurons in area V1 to sequences of patterns in which the similarity among patterns will be modulated in different ways. Mathematical models of dynamic aggregation, encoding, and reduction of information will be developed. The models will be augmented later to explore different possible roles that memory and attention can play in this aggregation of visual information. Finally, the models based on adaptive versus fixed cell responses will be compared. Of major interest throughout will be the investigation of perceptual capabilities that can emerge at the level of processing of the primary visual cortex and the innovative machine vision algorithms that can be developed that capture these capabilities. In addition to the potential of producing machine vision systems with capabilities of real vision, the work will contribute to the advancement of basic scientific understanding in the field of neurosciences, in particular of processes in the visual cortex. Neuromorphic system design is a developing new discipline that meshes neuroscience and mathematical modeling. Furthermore, the project will be a three way international collaboration between laboratories at the University of California at San Diego, the Algerian Ecole Nationale Polytechnique, and the Brain Science Institute in Japan. This will provide a rare opportunity for the students to work in an international research setting and to establish collaborative ties they may benefit from in the future in their independent research careers.

This material is based upon work supported by the National Science Foundation under Grant No. IIS-0082119.

Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.