Minchul Craig Shin

It's trivial though.

Are there any plans to update recent SOTA algorithms?

Pretrained weights trained on landmark dataset

3

I share the pretrained weight trained on landmarkd dataset. Please download it from the following url: https://drive.google.com/open?id=1dbdaDyVeIb53iGh4Uk5kA4in9-uURoLM

If there are any cool applications(but not widely known maybe..?) using GAN, let's share that trendy information! :1st_place_medal: please feel free to leave your comment here :metal: :metal: :metal:

fix python3 compatibility issue

1

+ fixed relative path issue. (https://github.com/primetang/pyflann/issues/1) + now, the code supports Python3. (tested on python3.4)

I figure `self.inv_c = np.sqrt(2.0/(in_ch))` should be ` self.inv_c = np.sqrt(2.0/(in_ch*ksize**2))` like EqualizedConv2d layer. What do you think? :0 original code: ~~~ class EqualizedDeconv2d(chainer.Chain): def __init__(self, in_ch, out_ch, ksize, stride,...

Has anyone succeeded in reproducing the results?

4

I am still struggling with training VQ-GAN in the first stage, not even the conditional transformer which is a second stage. The result looks fine before the discriminator loss is...

differences between cyclegan and discogan

3

(https://www.quora.com/What-is-the-difference-between-CycleGAN-and-DiscoGAN-They-both-seem-to-be-the-same-thing) + CycleGAN has a single cycle-consistency loss. Also, when comparing F(G(y)) with y, it seems that the two papers use different forms of distance measures (MSE, hinge-loss). + CycleGan...

It seems the network learns to ignore the encoded noise z_hat at the end of the training. Need to configure.

My PyTorch Implementation of DeLF

2

Hi, I found your repo, and just uploaded the code DeLF training code I implemented. (I've done this project quite long time ago, but was lazy enough to upload it...