BETHGE LAB · The Conjoint Effect of Divisive Normalization and Orientation Selectivity on Redundancy Reduction

Code Software Datasets Workshops & Tutorials

To reproduce the figures and Plots of the paper, download our code and results here, extract it and add all its directories to your matlab path.

If you want to train the models yourself, load the data from the data directory and call FmainV6 with it.

>> load data/Bristol_cropped_LMS.mat >> FmainV6(allImg)

Alternatively, you can load our learnt model from the results directory and reproduce the plots and tables with the following commands:

>> load results/ICA_data:BristolGray_patchsize:15_remDC:1_N=39936Ntest=49920V2.0_1-10.mat >> FgenerateFigure(res,{'HAD','SYM','ICA'},{'gN','RMixLogN'}) >> FplotRadialTransformation(res,{'HAD','SYM','ICA'})