GitHub - thu-fit/iGAN: Interactive Image Generation via Generative Adversarial Networks

Modified Code for Customized Training

Trained on Re:0. See GAN_Animate for detail.

[Project] [Youtube] [Paper]
A research prototype developed by UC Berkeley and Adobe CTL.

Name		Name	Last commit message	Last commit date
Latest commit History 47 Commits
GAN_Animate		GAN_Animate
datasets/scripts		datasets/scripts
lib		lib
model_def		model_def
models/scripts		models/scripts
pics		pics
res		res
train_dcgan		train_dcgan
ui		ui
.DS_Store		.DS_Store
.gitignore		.gitignore
LICENSE		LICENSE
README.md		README.md
constrained_opt.py		constrained_opt.py
constrained_opt_theano.py		constrained_opt_theano.py
download_res.sh		download_res.sh
generate_samples.py		generate_samples.py
iGAN_main.py		iGAN_main.py
iGAN_predict.py		iGAN_predict.py
iGAN_script.py		iGAN_script.py
run_main.sh		run_main.sh
run_script.sh		run_script.sh
theano概念.md		theano概念.md
update_server.sh		update_server.sh
解决theano安装问题.md		解决theano安装问题.md