segnet icon indicating copy to clipboard operation
segnet copied to clipboard

Published 20 hours ago • verified publisher icon preddy5

Feature Request: indices-based upsampling layer

Open waissbluth opened this issue 8 years ago • 5 comments

Hello, does anyone have an implementation of this, or any thoughts on how to implement it?

Thanks

waissbluth avatar Oct 28 '16 19:10 waissbluth

Hey @waissbluth I didnt get enough time to work on it yet, I'll probably work on in a week or two.

preddy5 avatar Oct 29 '16 12:10 preddy5

Hey, is there anything I can do to help out with this? would really appreciate the feature.

waissbluth avatar Dec 29 '16 20:12 waissbluth

Hey, I would also love this feature! :) Can I do anything to help?

Maybe a custom Maxpooling2D layer which produces an auxiliary output and an Upsample2D which would take an optionnal additionnal input would work?

mpariente avatar Mar 08 '17 15:03 mpariente

Hi all, this guy implemented a class he called Depool2D, I'm not sure if it's what you are looking for https://github.com/nanopony/keras-convautoencoder/blob/master/autoencoder_layers.py

redouanelg avatar Mar 16 '17 09:03 redouanelg

Thanks for the link ! I saw it already and I have troubles understanding his DependentDense. And also his DePool2D doesn't receive a switch from the MaxPooling2D if I read well.

mpariente avatar Mar 16 '17 09:03 mpariente