search

BVLC / caffe

Caffe: a fast open framework for deep learning.
http://caffe.berkeleyvision.org/
Other
34.12k stars 18.69k forks source link

How to implement VAE/DCGAN with upsampling (aka. backward convolution or fractional striding) in caffe? #5055

Closed darkAlert closed 7 years ago

darkAlert commented 7 years ago

How to implement VAE/DCGAN (decoder/generator) with upsampling layer (aka. backward convolution or fractional striding) in caffe?

I was able to implement the GAN (because only InnerProductLayer is needed). But I encountered a problem with upsampling operation while implementing decoder/generator net.

vra commented 7 years ago

Hi @darkAlert, can you share your solution to implement the GAN using caffe? Thanks!

zhangqianhui commented 7 years ago

@darkAlert Thanks for your sharing your solution to implement the gan ?

jeremy-rutman commented 7 years ago

There is a deconvolution layer available out of the box - did you get somewhere with dcgan?

shelhamer commented 7 years ago

Please discuss modeling and usage on the mailing list.

From https://github.com/BVLC/caffe/blob/master/CONTRIBUTING.md:

Please do not post usage, installation, or modeling questions, or other requests for help to Issues. Use the caffe-users list instead. This helps developers maintain a clear, uncluttered, and efficient view of the state of Caffe.