maxmustermann/animegan2-pytorch

Go to file

xhlulu f9b71ce9e5 Add weights from gdrive		2021-11-06 16:57:09 -04:00
samples	upsampling method changed + face model added	2021-02-21 15:04:52 +09:00
weights	Add weights from gdrive	2021-11-06 16:57:09 -04:00
.gitignore	Initial commit	2021-02-16 20:34:22 +09:00
convert_weights.py	typo fixed	2021-02-17 14:03:56 +09:00
hubconf.py	add face2paint	2021-11-06 16:51:14 -04:00
LICENSE	Create LICENSE	2021-08-23 07:45:19 +09:00
model.py	additional cli	2021-03-03 19:44:57 +09:00
README.md	Update README.md	2021-10-17 20:18:11 +09:00
test.py	additional cli	2021-03-03 19:44:57 +09:00
test_faces.ipynb	upsampling method changed + face model added	2021-02-21 15:04:52 +09:00

README.md

PyTorch Implementation of AnimeGANv2

Weight Conversion from the Original Repo (Requires TensorFlow 1.x)

git clone https://github.com/TachibanaYoshino/AnimeGANv2
python convert_weights.py

Inference

python test.py --input_dir [image_folder_path] --device [cpu/cuda]

Results from converted [Paprika] style model

(input image, original tensorflow result, pytorch result from left to right)

Note: Training code not included / Tested on RTX3090 + PyTorch1.7.1 / Results from converted weights slightly different due to the bilinear upsample issue

Additional Models

Webtoon Face [ckpt]

samples

Trained on 256x256 face images. Distilled from webtoon face model with L2 + VGG + GAN Loss and CelebA-HQ images. See test_faces.ipynb for details.

Face Portrait v1 [ckpt]

samples

Trained on 512x512 face images.

📺

Face Portrait v2 [ckpt]

samples

Trained on 512x512 face images. Compared to v1, 🔻beautify 🔺robustness

🦑 🎮 🔥