Multi-level-DCNet
Multi-level-DCNet copied to clipboard
primary capsule activation
Hi, In the paper you stated the following:
"We visualized what part of an image is activated by a 8D primary capsule of DCNet by guided back-propagation and noticed that every primary capsule is generated by a small spatial area of the input image"
How did you obtained the visualization of the activation? Could you provide more details please? Is the code for this visualization available? I would like to play around and see which parts of images activate which capsules.
Thank you very much for your time and help. Have a great day.