Interrupted traincascade output to xml
Is there a way, other than having to re-run traincascade with less stages specified than the last stage the process was interrupted at, to generate the desired xml file output that would usually happen once traincascade has completed all initial stages? I.e. I'm trying to avoid having to run a lengthy traincascade session of say 11 stages all over, to obtain an xml file from already completed 10 stages but no xml file yet.
Tks,
If in the traincascade arguments you specified it would have 11 stages, any .xml file taken from just 10 trained stages will be an incomplete classifier that will not work well at all.