BeautyGAN-pytorch

Unofficial implementation of ACM MM 2018 paper: "BeautyGAN: Instance-level Facial Makeup Transfer with Deep Generative Adversarial Network"

it can successfully run.

downing the 'vgg_conv.pth' : https://bethgelab.org/media/uploads/pytorch_models/vgg_conv.pth

put it at ./addings/
downing dataset. Dataset can be found in project page: http://colalab.org/projects/BeautyGAN, containing of all/images/ and all/segs

put all/images/* and all/segs/* into ./data/makeup/makeup_final/
using generate.py to generate '*.txt' eg. 'train_MAKEMIX.txt.

put four txt files into ./data/makeup/makeup_final/
python train.py

Name		Name	Last commit message	Last commit date
Latest commit History 6 Commits
data_loaders		data_loaders
ops		ops
README.md		README.md
config.py		config.py
dataloder.py		dataloder.py
generator.py		generator.py
net.py		net.py
solver_cycle.py		solver_cycle.py
solver_makeup.py		solver_makeup.py
test.py		test.py
train.py		train.py
visualize.py		visualize.py