Modeling Large Cortical Networks With Growing Self-Organizing Maps (2002)
Self-organizing computational models with specific intracortical connections can explain many features of visual cortex. However, due to their computation and memory requirements, it is difficult to use such detailed models to study large-scale object segmentation and recognition. This paper describes GLISSOM, a method for scaling a small RF-LISSOM model network into a larger one during self-organization, dramatically reducing time and memory needs while obtaining equivalent results. With GLISSOM it should be possible to simulate all of human V1 at the single-column level using existing supercomputers. The scaling equations GLISSOM uses also allow comparison of biological maps and parameters between individuals and species with different brain region sizes.
In Computational Neuroscience, 44--46, 315-321, 2002.

James A. Bednar Postdoctoral Alumni jbednar [at] inf ed ac uk
Amol Kelkar Masters Alumni
Risto Miikkulainen Faculty risto [at] cs utexas edu

The LISSOM package contains the C++, Python, and Scheme source code and examples for training and testing firing-rate...