zmyvisions
commited on
Commit
•
0aa998f
1
Parent(s):
d162ca9
Update README.md
Browse files
README.md
CHANGED
@@ -1,47 +1,43 @@
|
|
1 |
-
---
|
2 |
-
|
3 |
-
|
4 |
-
|
5 |
-
|
6 |
-
|
7 |
-
|
8 |
-
|
9 |
-
|
10 |
-
|
11 |
-
|
12 |
-
|
13 |
-
|
14 |
-
|
15 |
-
|
16 |
-
|
17 |
-
|
18 |
-
|
19 |
-
|
20 |
-
|
21 |
-
|
22 |
-
|
23 |
-
|
24 |
-
|
25 |
-
|
26 |
-
- [
|
27 |
-
-
|
28 |
-
|
29 |
-
|
30 |
-
|
31 |
-
- [
|
32 |
-
-
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
|
37 |
-
-
|
38 |
-
|
39 |
-
|
40 |
-
|
41 |
-
|
42 |
-
-
|
43 |
-
- Spaces/Models re-packaging and implementation by [Shō Akiyama](https://github.com/Yijunmaverick/).
|
44 |
-
|
45 |
-
##### Special Thanks
|
46 |
-
- [Nima Boscarino](https://github.com/NimaBoscarino)
|
47 |
- [Omar Sanseviero](https://github.com/osanseviero)
|
|
|
1 |
+
---
|
2 |
+
{}
|
3 |
+
---
|
4 |
+
|
5 |
+
# AnimeBackgroundGAN (CartoonGAN by Chen et. al.)
|
6 |
+
|
7 |
+
<img src="https://m.media-amazon.com/images/M/MV5BZTExN2EwMmYtNDcwZS00ZmI1LTk1NGQtNTQ3YWFjMmY3YjQwXkEyXkFqcGdeQXVyNTU1OTUzNDg@._V1_.jpg" alt="5 Centimetres per Second directed by Makoto Shinkai" style="height: 300px;"/>
|
8 |
+
|
9 |
+
- [Makoto Shinkai (新海誠)](https://en.wikipedia.org/wiki/Makoto_Shinkai) pre-trained model from [CartoonGAN](http://openaccess.thecvf.com/content_cvpr_2018/CameraReady/2205.pdf) `[Chen et al., CVPR18]`.
|
10 |
+
- This model can transform real-life photos into Japanese-animation-like backgrounds, following the style of movies such as [Kimi no Na wa](https://en.wikipedia.org/wiki/Kimi_no_Na_wa) with a photorealistic painting style.
|
11 |
+
- The implementation is in PyTorch (see [source code here](https://huggingface.co/spaces/akiyamasho/AnimeBackgroundGAN/blob/main/network/Transformer.py)).
|
12 |
+
- Check out the demo here:
|
13 |
+
|
14 |
+
[![Demo in Hugging Face Spaces](https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Spaces-blue)](https://huggingface.co/spaces/akiyamasho/AnimeBackgroundGAN)
|
15 |
+
|
16 |
+
# Other pre-trained model versions
|
17 |
+
|
18 |
+
The other versions were also trained from movies of the different Japanese animation directors.
|
19 |
+
|
20 |
+
##### Mamoru Hosoda(細田守)
|
21 |
+
- director of [Wolf Children](https://en.wikipedia.org/wiki/Wolf_Children), with a distinct mild and cool background style
|
22 |
+
- [Director Profile](https://en.wikipedia.org/wiki/Mamoru_Hosoda)
|
23 |
+
- **Model Repository**: https://huggingface.co/akiyamasho/AnimeBackgroundGAN-Hosoda
|
24 |
+
|
25 |
+
##### Satoshi Kon(今敏)
|
26 |
+
- director of [Paprika](https://en.wikipedia.org/wiki/Paprika_(2006_film)) with a distinct high contrast, reddish hue style
|
27 |
+
- [Director Profile](https://en.wikipedia.org/wiki/Satoshi_Kon)
|
28 |
+
- **Model Repository**: https://huggingface.co/akiyamasho/AnimeBackgroundGAN-Kon
|
29 |
+
|
30 |
+
##### Hayao Miyazaki(宮崎駿)
|
31 |
+
- director of [Howl's Moving Castle](https://en.wikipedia.org/wiki/Howl%27s_Moving_Castle_(film)) with a relatively soft and painterly style
|
32 |
+
- [Director Profile](https://en.wikipedia.org/wiki/Hayao_Miyazaki)
|
33 |
+
- **Model Repository**: https://huggingface.co/akiyamasho/AnimeBackgroundGAN-Miyazaki
|
34 |
+
|
35 |
+
### Credits
|
36 |
+
|
37 |
+
- Paper at [CartoonGAN: Generative Adversarial Networks for Photo Cartoonization](http://openaccess.thecvf.com/content_cvpr_2018/CameraReady/2205.pdf) `[Chen et al., CVPR18]`
|
38 |
+
- Original PyTorch implementation was created by [Yijun Li](https://github.com/Yijunmaverick/)
|
39 |
+
- Spaces/Models re-packaging and implementation by [Shō Akiyama](https://github.com/Yijunmaverick/).
|
40 |
+
|
41 |
+
##### Special Thanks
|
42 |
+
- [Nima Boscarino](https://github.com/NimaBoscarino)
|
|
|
|
|
|
|
|
|
43 |
- [Omar Sanseviero](https://github.com/osanseviero)
|