RobotJelly commited on
Commit
b6920a8
1 Parent(s): 456f536
Files changed (1) hide show
  1. README.md +12 -5
README.md CHANGED
@@ -6,19 +6,22 @@ tags:
6
  - GAN
7
  - spatially-adaptive normalization
8
  - Encoder
 
9
  ---
10
 
11
  ## Model description
12
 
13
- More information needed
14
 
15
- ## Intended uses & limitations
16
 
17
- More information needed
 
 
18
 
19
  ## Training and evaluation data
20
 
21
- More information needed
22
 
23
  ## Training procedure
24
 
@@ -37,4 +40,8 @@ The following hyperparameters were used during training:
37
 
38
  ![Model Image](./model.png)
39
 
40
- </details>
 
 
 
 
6
  - GAN
7
  - spatially-adaptive normalization
8
  - Encoder
9
+ - Segmentation-maps
10
  ---
11
 
12
  ## Model description
13
 
14
+ In this, GauGAN architecture has been implemented for conditional image generation which was proposed in [Semantic Image Synthesis with Spatially-Adaptive Normalization](https://arxiv.org/abs/1903.07291).
15
 
16
+ GauGAN uses a `Generative Adversarial Network (GAN)` to generate realistic images that are conditioned on cue images and segmentation maps.
17
 
18
+ This repo contains the model for the notebook [**GauGAN for conditional image generation**](https://keras.io/examples/generative/gaugan/)
19
+
20
+ Full credits go to [Soumik Rakshit](https://github.com/soumik12345) & [Sayak Paul](https://twitter.com/RisingSayak)
21
 
22
  ## Training and evaluation data
23
 
24
+ Here, the [Facades dataset](https://cmp.felk.cvut.cz/~tylecr1/facade/) is used for training GauGAN model. Some custom layers that were added into the model are - SPADE (SPatially-Adaptive (DE) normalization), Residual block including SPADE & Gaussian sampler. Also, the GauGAN encoder consists of a few downsampling blocks. It outputs the mean and variance of a distribution as shown in this [image](https://i.imgur.com/JgAv1EW.png).
25
 
26
  ## Training procedure
27
 
40
 
41
  ![Model Image](./model.png)
42
 
43
+ </details>
44
+
45
+ <center>
46
+ Model Reproduced By <u><a href="https://github.com/robotjellyzone"><b>Kavya Bisht</b></a></u>
47
+ </center>