Dataset: Anime GAN Lite | Kaggle
I manually delete the desktop.ini file
this project is a simple implementation of DC-GAN to generate pictures by random noise.
How to use this model
- configure the model you want to use in the
python generate.pyin the terminal
How to train your own model
- put the dataset you wish to generate under
- configure the
config.pychange the argument
--datasetto your own dataset path
python train.pyin the terminal
:warning:Note: since normalization operation included, the color style between generate pic and original pic may not same.
- The output of the discriminator can be optimized, which may improve the network's performance
- It still has mode collapse problems in a given model. I will fix it later.