How to adapt Image Category Classification Using Deep Learning example with Fast R-CNN ?

Hi. I'm a novice in Matlab and Deep learning.

I already made a classification using the method used in this example "Image Category Classification" for my own dataset (700images in 7categories : 100 each) with the alexnet layers.

But now I want to use that classification to detect the objects in a bigger image using Faster R-CNN. I already made the RCNN training, but it was not accurate at all and took me 2 days for the training !

Any help ?