Get feature-map of a given region in the image from the ground truth

[praisethemoon]

Hello!

Do you have any idea on how can I get the feature map from the last layer of a given region (from ground truth) after passing the entire image throughout the RPN?

In particular:

# get the feature maps and output from the RPN
[Y1, Y2, F] = model_rpn.predict(X)

I am assuming that F is the feature map, and I want to get the feature-map of a small region in the image (pedestrian from GT) for i want to use that feature map for training a SVM classifier.

Cheers!