This source code implements the microcircuit model of attention, described in the article: Beuth F, Hamker FH (2015) "A mechanistic cortical microcircuit of attention
for amplification, normalization and suppression", Vision Research. Please cite this paper if this code is used in or to motivate any publications.
In the code, L denotes number of feature in layer 4 and layer 2/3, and X the number of locations in layer 4. Layer 2/3 has less space, denoted by Xpooled, due to spatial pooling.
Getting started
The files calculateModel.m and initModel.m implement the model, whereby the former executes the calculations and the later creates a structure (called 'model object') containing all parameters. The parameters are passed to calculateModel(..) via the model object.
Parameters should be modified in the user function, hence outside of initModel.m. A description of all parameter can be found in the file initModel.m.
Thus, a user function should call the files typically in this order:
1) Create model structure: objModel = initModel(..)
2) Modify parameters, e.g. objModel.vSP = 5
3) Run model: calculateModel(.., objModel)
The files figure*.m reproduce the figures of the publication. They can be used as examples.
The file calculateModel.m can also be executed standalone as a toy-example. The example simulates two stimuli with spatial attention to the second one. They do not suppress each other, because they are not placed within a receptive field, so feature-based suppression is not occurring. They are different, so that surround suppression is also not occurring.
Generated on Tue Apr 21 2015 17:28:22 for Microcircuit_Model_of_Attention by 1.8.1.2