-
I am wondering how can I evaluate the model by a real image instead of a generated image by StyleGAN.
The input image usually should be embed into the latent space of GAN by a latent optimizer to …
-
Hi! Thanks for the very neat training code!
In the generator training part (line 142 of train.py). Shouldn't you be updating the parameters of encoder network instead of the decoder (referred to as…
-
## Abstract
- Proposes Deliberation Network, a two-pass encoder-decoder framework with second-pass conducting deliberation
- SoTA on WMT 2014 En-Fr NMT task
## Details
- ```Deliberation``` is a …
-
Read the papers in [Deep Learning’s Most Important Ideas](https://www.kdnuggets.com/2020/09/deep-learnings-most-important-ideas.html).
- [x] Tackling ImageNet with AlexNet and Dropout
- [x] [I…
-
## 0. 論文
[Unsupervised Neural Machine Translation with Weight Sharing](https://arxiv.org/abs/1804.09057v1)
Zhen Yang, Wei Chen, Feng Wang, Bo Xu
## 1. どんなもの？
潜在空間のパラメータを共有して行う教師なしニューラル機械翻訳
##…
-
As far as I understand the Pixel Discriminator is equivalent to a PatchGAN of size 1x1, meaning the decoder tries to decide if each pixel is real or fake. But without taking into account surrounding p…
-
## 一言でいうと
GANのようにピクセル単位の画像(ラスタライズ)ではなく、ストローク単位の画像(ベクター画像)を生成する試み(人が絵を描くときは後者に近い)。
ストロークをペンの状態で表現し(前座標との差分・始点・終点・描画終了)、これをRNN(Encoder/Decoder)+VAEでモデル化
The above is the generative module in the paper.
Can I ask why you applied the x'or…
-
Additionally, I want to ask why some losses increase during training. Should I choose the last checkpoint or the best checkpoint?