The corrective learning code is based on Hongzhi Wang's segAdapter release 1.9

There is another branch of the code in ANTS, but I am not sure of its state relative to Hongzhi's code
