Currently the InputDual module simply pairs the 
-th images from each
of its left and right Input modules together. A desirable extension would be
to have InputDual examine the identifier fields in the StereoImages it
receives, and then pair stereo images together based on this. This would most
notably require handling the case where frames are delayed, out-of-order or
simply missing, presumably with some form of caching or interpolation scheme.