An oscillatory neural model of multiple object tracking (Kazanovich and Borisyuk 2006)

 Download zip file 
Help downloading and running models
Accession:79145
An oscillatory neural network model of multiple object tracking is described. The model works with a set of identical visual objects moving around the screen. At the initial stage, the model selects into the focus of attention a subset of objects initially marked as targets. Other objects are used as distractors. The model aims to preserve the initial separation between targets and distractors while objects are moving. This is achieved by a proper interplay of synchronizing and desynchronizing interactions in a multilayer network, where each layer is responsible for tracking a single target. The results of the model simulation are presented and compared with experimental data. In agreement with experimental evidence, simulations with a larger number of targets have shown higher error rates. Also, the functioning of the model in the case of temporarily overlapping objects is presented.
Reference:
1 . Kazanovich Y, Borisyuk R (2006) An oscillatory neural model of multiple object tracking. Neural Comput 18:1413-40 [PubMed]
Model Information (Click on a link to find other models with that property)
Model Type: Connectionist Network;
Brain Region(s)/Organism:
Cell Type(s):
Channel(s):
Gap Junctions:
Receptor(s):
Gene(s):
Transmitter(s):
Simulation Environment: C or C++ program;
Model Concept(s): Oscillations; Spatio-temporal Activity Patterns; Simplified Models;
Implementer(s): Kazanovich, Yakov [yakov_k at impb.psn.ru]; Borisyuk, Roman [rborisyuk at plymouth.ac.uk]; 
// Parset.cpp - Network parameters and initial state of variables definition.
// Project MOT
// Kazanovich June 2005

#include "stdafx.h"
#include "mot.h"

extern struct parameters par;
extern struct integration integr;
extern struct imageproc improc;
extern struct image im;

double rn(void);
//=================== s/p Pardef ===================
// Network parameters definition

const double PI = 3.1415926535;

void Pardef()
{							
par.n = nrows*ncolumns;

// CO
//par.camp = 8.0;
par.camp = 10.0;			// initial amplitude of the CO
par.comega = 6.0;		// initial natural frequency of CO
par.cteta =  0.;		// initial phase of CO

// POs
par.amp = 2.0;			// initial amplitude of POs

// so it was par.COtoPOw = -6.;		// connection weights
// so it was par.POtoCOw = 2.5;			
par.COtoPOw = -7.;		// connection weights
//par.COtoPOw = -15.;
par.POtoCOw = 5;			
par.POtoPOlocw = 40.0;	// weight of local interaction
//par.COtoCOw = -3.6;		// weight of interaction between COs 
par.COtoCOw = -7.0;
par.COtoCOw_expos = -12.0;	// weight of interaction between COs during exposition
//par.COtoCOw_expos = -15.0;
par.POtoPOcolw = 40.0;	// weight of interaction betwee POs in the same column

// Image parameters
im.r = 3;

// Integration parameters
integr.eps = 0.000001;	// Runge-Kutta integration parameters
integr.h1 = 0.01;
integr.hmin = 1.e-15;
integr.dt = 0.1;

integr.maxt = 72.0;		// Time of system functioning
integr.expos = 6.0,		// Initial exposition without any changes of an image 
integr.probe = 5.0;		// Testing period
integr.move_type = without_intersection;	// Type of movement
integr.movement_dt = 5; // Time interval between movements
// Parameters of eq.

// Parameters of function g in CO's eq
integr.gs1 = 0.1;		
integr.gq1 = 1.0;
integr.gs2 = 0.2;
integr.gq2 = 0.6;
integr.gs3 = 2*PI;

integr.ga1 = integr.gq1/integr.gs1;
integr.ga2 = (integr.gq1 - integr.gq2)/(integr.gs1 - integr.gs2);
integr.gb2 = (integr.gq2*integr.gs1 - integr.gq1*integr.gs2)/
			(integr.gs1 - integr.gs2);
integr.ga3 = integr.gq2/(integr.gs2 - integr.gs3);
integr.gb3 = - integr.gq2*integr.gs3/(integr.gs2 - integr.gs3);

// parameters for amplitude equation
// max amplitude = gamma(1 + dzeta)
integr.beta1 = 1.0;				
integr.beta2 = 1.0;		// !?! 2.0
integr.beta_CO = 1.0;
integr.gamma = 10.0; 
integr.ksi = 0.9;
integr.eta = 0.02;	// 0.015 !?!
integr.dzeta = 0.1;
integr.gamma1 = 1.7;	
integr.dzeta1 = 1.0;

// parameter for natural frequency equation
integr.alpha = 2.0;

// Resonance threshold
integr.resthresh = 0.8*integr.gamma*(1 + integr.dzeta);
// min normalization parameter in the eq for the CO
integr.nresmin = (im.r*2 + 1)*(im.r*2 + 1);		
// parameters for duration of activity states
integr.noise = 3.0;

// Parameters to controle the strength connection from POs to CO
improc.intenstarget = 5.0;	// the value of intensity for a target
improc.intensdistract = 5.0;// the value of intensity for a distractor
//improc.highsaliency = 5.0;
improc.highsaliency = 5.0;
improc.lowsaliency = 0.2;
improc.normalsaliency = 1.0;
}

Loading data, please wait...