1 | initial version |
I would start reading this forum for a starter. At the start page type in traincascade in the search option. There are some pretty nice guidelines out there. For example, read this and this to understand the training parameters better.
For the actual implementation, there is a topic here somewhere with the response of the original author, lemme see if I can find it and I will get back to you!
Also reading through the Viola & Jones paper, seems logical to me, to get the grasp of the actual algorithm.
2 | No.2 Revision |
I would start reading this forum for a starter. At the start page type in traincascade in the search option. There are some pretty nice guidelines out there. For example, read this and this to understand the training parameters better.
For the actual implementation, there is a topic here somewhere with the response of the original author, lemme see if I can find it and I will get back to you!
Also reading through the Viola & Jones paper, seems logical to me, to get the grasp of the actual algorithm.
EDIT 1:
I have looked up some extra information. Basically searching for traincascade and @Maria Dimashova gives you tons of good topics with alot of information, since she is the one who created the documentation and the algorithm.
Click here for more info.