capsule-networks icon indicating copy to clipboard operation
capsule-networks copied to clipboard

Published 20 hours ago • verified publisher icon gram-ai

Can the 8 32-channel conv be replaced with a single 256-channel conv in PrimaryCapsule?

Open njzwj opened this issue 6 years ago • 1 comments

Thanks for sharing the nicely written code.

I wonder that is there any difference between using 8 32-channel convlayer and a single 256-channel convlayer? It seems like the "In total PrimaryCapsules has [32*6*6] capsule outputs (each output is an 8D vector)" mentioned in Dynamic Routing Between Capsules can be simply implemented as a 256-channel convlayer followed by a reshaping operation.

But I haven't seen anyone using 256-channel conv so far. Am I thinking wrong?

njzwj avatar Jul 09 '19 13:07 njzwj

I have used such an implementation for the PrimaryCaps layer in my repository (https://github.com/gchochla/caps/blob/master/caps/layers.py#L33). I had to use an extra variable to separate them for the output of the layer though.

gchochla avatar Feb 13 '20 14:02 gchochla