Unsupervised Learning of Overlapping Image Components Using Divisive Input Modulation

Figure 2

(a) A simple neural network of the type used
by all the algorithms described in this article (the symbols are the same as
those used in Figure 1). This network has one output node which receives equal
strength weights from two error-detecting nodes (i.e., ). Each
error-detecting node receives equal strength input from two-image pixels (i.e., ). Each
subfigure in (b) and (c) shows the steady-state activation strength of the
output node and the two error-detecting nodes in this simple network calculated
using (b) the sequential NMF algorithm, and (c) the divisive input modulation
algorithm. The steady-state responses are calculated for different weight
values (indicated by the width of each connection which is proportional to its
strength). From top to bottom in (b) and (c) the weights (i.e., and ) are equal to
0.25, 0.5, 1, and 2. Note that there is no stochastic element in the
calculation of the neural responses generated by these algorithms, so identical
results will be generated each time the network is simulated with these weight
values.

(a)

(b)

(c)

Article of the Year Award: Outstanding research contributions of 2020, as selected by our Chief Editors. Read the winning articles.