Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

deconv vs. conv2d_transpose #2

Closed
kchen92 opened this issue Jan 29, 2016 · 2 comments
Closed

deconv vs. conv2d_transpose #2

kchen92 opened this issue Jan 29, 2016 · 2 comments

Comments

@kchen92
Copy link
Contributor

kchen92 commented Jan 29, 2016

It seems to me that you are using the tensorflow deconv2d which performs a transpose convolution (see conv2d_transpose at https://github.com/tensorflow/tensorflow/blob/4c3b060b9b063a247172c939e0e1314577b581e3/tensorflow/python/ops/nn_ops.py). However, the deconv that we want to perform is an unpooling followed by convolution. Could you clarify whether my understanding is correct? Thanks.

Also relevant: tensorflow/tensorflow#256

@ikostrikov
Copy link
Owner

Yes. That's correct. In some papers it's wrongly called deconvolution so I decided to keep it this way.

For example, they mention it here (Figure 1):
http://arxiv.org/pdf/1511.06434v2.pdf

@kchen92
Copy link
Contributor Author

kchen92 commented Jan 29, 2016

Ok. I have to do some reading still but I think I get it. Thanks.

@kchen92 kchen92 closed this as completed Jan 29, 2016
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants