Hi, I want to train a mask rcnn model on a custom dataset.
Will be using the Detectron repo for this.
Some objects in the dataset are very small
by up scaling all images and by using the following RPN
RPN_ASPECT_RATIOS: (0.25, 0.5, 1.0, 2.0, 4.0)
Hopefully the small objects can be detected
The question is, should I edit the params for ROI pooling?
I would be referring to
Thanks for your advice!