initializeWTA - intitializes a winner-take-all network. wta = initializeWTA(saliencyMap,saliencyParams) Initializes a winner-take-all network of leaky integrate and fire neurons with the current inputs to the neurons set proportional to the values of the saliencyMap. See also evolveWTA, defaultLeakyIntFire, runSaliency, defaultSaliencyParams, dataStructures.
0001 % initializeWTA - intitializes a winner-take-all network. 0002 % 0003 % wta = initializeWTA(saliencyMap,saliencyParams) 0004 % Initializes a winner-take-all network of leaky 0005 % integrate and fire neurons with the current 0006 % inputs to the neurons set proportional to the 0007 % values of the saliencyMap. 0008 % 0009 % See also evolveWTA, defaultLeakyIntFire, runSaliency, 0010 % defaultSaliencyParams, dataStructures. 0011 0012 % This file is part of the SaliencyToolbox - Copyright (C) 2006-2007 0013 % by Dirk B. Walther and the California Institute of Technology. 0014 % See the enclosed LICENSE.TXT document for the license agreement. 0015 % More information about this project is available at: 0016 % http://www.saliencytoolbox.net 0017 0018 function wta = initializeWTA(salmap,salParams) 0019 0020 wta.sm = defaultLeakyIntFire; 0021 wta.sm.C = 5e-8; 0022 wta.sm.Einh = 0; 0023 wta.sm.Eexc = 0; 0024 wta.sm.Gleak = 1e-7; 0025 wta.sm.Ginh = zeros(size(salmap.data)); 0026 wta.sm.GinhDecay = 0.9999; 0027 wta.sm.DoesFire = 0; 0028 wta.sm.I = salmap.data * salParams.smOutputRange + ... 0029 salParams.noiseAmpl * rand(size(salmap.data)) + ... 0030 salParams.noiseConst; 0031 debugMsg('salmap input into WTA',wta.sm.I); 0032 wta.sm.V = zeros(size(salmap.data)); 0033 0034 wta.exc = defaultLeakyIntFire; 0035 wta.exc.I = zeros(size(salmap.data)); 0036 wta.exc.V = zeros(size(salmap.data)); 0037 wta.exc.Ginh = 1e-2; 0038 0039 wta.inhib = defaultLeakyIntFire;