%% cd '/Volumes/VOLUME2_MAC_INT_DH/D_DRIVE/CURRENT_WORK/PROMET_2009_MATLAB' % #K % #G % DIMENSION_FREQMAP: 30.0; % DIMENSION_AUDMAP: 4.0; % DIMENSION_EYEPOS: 10.0; % DIMENSION_HIDDEN: 10.0; % DIMENSION_OUTPUT: 20.0; % % SIGMA_FREQMAP: 0.5; % SIGMA_MOTMAP: 10.0; % % GAIN_FREQMAP: 1; % EXP_FREQMAP: 40; % % GAIN_EYEPOS: 1; % GAIN_MOTOR: .7; % % BIAS_HID: 5; % % LEARNING_RATE: 0.5; % % AZIMUTH: 0.0; % EYEPOSITION: 0.0; % FREQUENCY: 4950.0; % INTENSITY: 0.0; % % RANDOM_FLAG: 1; % CONTINUE_FLAG: 0; % % MOTORERROR_FLAG: 0; % AZIMUTH_FLAG: 0; % EYE_FLAG: 0; % FREQUENCY_FLAG: 0; % INTENSITY_FLAG: 0; % % DEMO_FLAG: 1000; % DEMO_WAIT: 100; % STORE_FLAG: 1; % STORE: 0; % % WEIGHTS_RANGE: .5; % % AZIMUTH_RANGE: 60.0; % EYEPOS_RANGE: 40.0; % MOTORERROR_RANGE: 100.0: % LOW_FREQUENCY_RANGE: 100.0; % HIGH_FREQUENCY_RANGE: 10000.0; % INTENSITY_RANGE: 0.40; % % #V % 1/1.wht; % 2/2.wht; % 3/3.wht; % 4/4.wht; % 5/4.wht; % #E % % % % /* *************** SETTING DIMENSION VALUES ************** */ % % Nfreq = (unsigned short)get_bat(bt, 0, "DIMENSION_FREQMAP", &Found); % Naud = (unsigned short)get_bat(bt, 0, "DIMENSION_AUDMAP", &Found); % Nmsc = (unsigned short)get_bat(bt, 0, "DIMENSION_EYEPOS", &Found); % Nmot = (unsigned short)get_bat(bt, 0, "DIMENSION_OUTPUT", &Found); % Nhidden = (unsigned short)get_bat(bt, 0, "DIMENSION_HIDDEN", &Found); % % % Sigma = get_bat(bt, 0, "SIGMA_MOTMAP", &Found); % Sig_freq= get_bat(bt, 0, "SIGMA_FREQMAP", &Found); % % Gain_freq= get_bat(bt, 0, "GAIN_FREQMAP", &Found); % Gain_pos = get_bat(bt, 0, "GAIN_EYEPOS", &Found); % Gain_mot = get_bat(bt, 0, "GAIN_MOTOR", &Found); % Exp_freq = get_bat(bt, 0, "EXP_FREQMAP", &Found); % % Eye_range = get_bat(bt, 0, "EYEPOS_RANGE", &Found); % Mot_range = get_bat(bt, 0, "MOTORERROR_RANGE", &Found); % Azi_range = get_bat(bt, 0, "AZIMUTH_RANGE", &Found); % Freq_min = get_bat(bt, 0, "LOW_FREQUENCY_RANGE", &Found); % Freq_max = get_bat(bt, 0, "HIGH_FREQUENCY_RANGE", &Found); % Int_range = get_bat(bt, 0, "INTENSITY_RANGE", &Found); % % % % LEARNING_RATE = get_bat(bt, 0, "LEARNING_RATE", &Found); % BIAS_HID = get_bat(bt, 0, "BIAS_HID", &Found); % Weights_range = get_bat(bt, 0, "WEIGHTS_RANGE", &Found); % % MOTORERROR_FLAG = (short)get_bat(bt, 0, "MOTORERROR_FLAG", &Found); % AZIMUTH_FLAG = (short)get_bat(bt, 0, "AZIMUTH_FLAG", &Found); % EYE_FLAG = (short)get_bat(bt, 0, "EYE_FLAG", &Found); % FREQUENCY_FLAG = (short)get_bat(bt, 0, "FREQUENCY_FLAG", &Found); % INTENSITY_FLAG = (short)get_bat(bt, 0, "INTENSITY_FLAG", &Found); % RAND_FLAG = (short)get_bat(bt, 0, "RANDOM_FLAG", &Found); % % CONTINUE_FLAG = (short)get_bat(bt, 0, "CONTINUE_FLAG", &Found); % STORE_FLAG = (short)get_bat(bt, 0, "STORE_FLAG", &Found); % DEMO_FLAG = (short)get_bat(bt, 0, "DEMO_FLAG", &Found); % STORE = (unsigned long)get_bat(bt, 0, "STORE", &Found); % WAIT = (unsigned long)get_bat(bt, 0, "DEMO_WAIT", &Found); % % % /* ************* CREATING VECTORS AND MATRICES *********** */ % % Audmap = vector(1, Naud); % Eyepos = matrix(1, 2, 1, Nmsc); % Cochlea = matrix(1, 2, 1, Nfreq); % % Hidden = vector(1, Nhidden); % % MotorError = vector(1, Nmot); % Teacher = vector(1, Nmot); % Error_hm = vector(1, Nmot); % Error_ah = vector(1, Nhidden); % % Bias_HID = vector(1, Nhidden); % Bias_MOT = vector(1, Nmot); % Bias_AUD = vector(1, Naud); % % Weights_AH = matrix(1, Nhidden, 1, Naud); % Weights_CA = matrix(1, Naud, 1, 2 * Nfreq); % Weights_EH = matrix(1, Nhidden, 1, 2 * Nmsc); % Weights_HM = matrix(1, Nmot, 1, Nhidden); % % /* ************************************************************************************************* */ % %% % DIMENSION_FREQMAP: 30.0; % DIMENSION_AUDMAP: 4.0; % DIMENSION_EYEPOS: 10.0; % DIMENSION_HIDDEN: 10.0; % DIMENSION_OUTPUT: 20.0; % % SIGMA_FREQMAP: 0.5; % SIGMA_MOTMAP: 10.0; % % GAIN_FREQMAP: 1; % EXP_FREQMAP: 40; % % GAIN_EYEPOS: 1; % GAIN_MOTOR: .7; % % BIAS_HID: 5; % % LEARNING_RATE: 0.5; %% % /* ******************************* CALC COCHLEA-LAYER ACTIVATION VALUES ****************** */ % /* % This function calculates the activity values of the cochlea layers (left and right) which are presented in a forward % direction to the audmap layer % */ % % /* % function that generates Gaussian receptive field profile: % F = exp( -[(x-x0)^2]/[2sigma^2] ) % */ %void coch_input(cochlea, azi, nstimuli, Is, freq_flag) clc freq_flag=0; Nfreq=30; Freq_min=100; Freq_max=10000; Gain_freq = 1.0; Exp_freq = 40.0; Sig_freq = 0.5; azi=30; nstimuli=4; Is=1; %% azi=-90:1:90; IL=Is-exp( +(1/Exp_freq)*(azi - 90)) IR=Is-exp( -(1/Exp_freq)*(azi + 90)) figure(1) subplot(211) plot(azi,IL) subplot(212) plot(azi,IR) %% clc cf=1000 azi=-60; Is=1.0; Gain_freq = 1.0; Exp_freq = 40.0; Sig_freq = 4; Nfreq=100; nstimuli=10; freq_flag=0; Nfreq=100; Freq_min=100; Freq_max=10000; % IL=Is-exp( +(1/Exp_freq)*(azi - 90)); % if(IL<0) IL=0.0; end; % IR=Is-exp( -(1/Exp_freq)*(azi + 90)); % if(IR<0) IR=0.0; end; figure(1) hold off cochlea=zeros(2,Nfreq); for k=1:1:nstimuli %cf = randi(10000,1,1)+99; %io = (log(cf)-log(Freq_min)) / (log(Freq_max)-log(Freq_min)) * (Nfreq -1) +1 io = randi(Nfreq,1,1) m = (Nfreq - 1) / (log(Freq_max) - log(Freq_min)); logf = ((io - 1.0)/m) + log(Freq_min); cf = exp(logf); for i=1:1:Nfreq % changed dependency blocking is freqency dependent % % High freq much diminution % % Low frequency less diminution ((float)Exp_freq+i*20) % IL=Is-exp( +(1/(Exp_freq + i*20))*(azi - 90)); if(IL<0) IL=0.0; end; IR=Is-exp( -(1/(Exp_freq + 1*20))*(azi + 90)); if(IR<0) IR=0.0; end; rate_r = Gain_freq * IR * exp( -((i-io)*(i-io)) / (2*(Sig_freq*Sig_freq)) ); if (rate_r 90.0) rate_l = 0.0; else rate_l = Is + (Gain_freq * (1 - exp( +(azi - 90.0) / Exp_freq) )); end if(rate_r<0) rate_r=0;end if(rate_l<0) rate_l=0;end arg = ( sqrt(i-i0) / (2 * sqrt(Sig_freq)) ); cochlea(1,i) = cochlea(1,i) + rate_r * exp(-arg) if (cochlea(1,i) > rate_r) cochlea(1,i) = rate_r; end cochlea(2,i) = cochlea(2,i) + rate_l * exp(-arg) if (cochlea(2,i) > rate_l) cochlea(2,i) = rate_l; end end end % /* end k loop */ %/* *************** END COCHLEA OUTPUT *************** */ figure(1) plot((cochlea(1,:))) %%