How to train the object detector

[alphaics-prateek]

Hi burak and chris,

I wanted to know how do we need to train the object detector?

Do we break images into patches and feed them to the coarse net? do we need to label the patches separately to identify whether images are in the patches or not? and are you resizing the 64x64 patch to 320x320 for training yolo?

Also about storing the output in the desired format, can you please share the code to take yolo o/p and store it in the folder pattern you have described in the readme?

[bling-bling-only]

I also want to know.Do you know the answer