JiaRenChang/Batch_Normalized_Maxout_NIN

module usage

albanie opened this issue · 2 comments

This is a really useful repo. I was thinking adding your maxout implementation to a matconvnet vl_contrib module to make it easier to find for others, would you be OK with that?

It's OK. I am so glad!

Fantastic. I've pulled things together into a module with a few minor changes to speed up the CUDA. The module is here: https://github.com/albanie/mcnMaxout. I've also added you as a collaborator in case you find anything you'd like to change. When I get the chance I'll try to also add in your funky NiN architecture :)