IE-GAN[arXiv]

Codes for paper "IE-GAN: An Improved Evolutionary Generative Adversarial Network Using a New Fitness Function and a Generic Crossover Operator"

Prerequisites

• Linux or macOS
• Python 3
• CPU or NVIDIA GPU + CUDA CuDNN

Getting Started

Preparation

• Preparing .npz files for Pytorch Inception metrics evaluation (cifar10 as an example):

python metric/inception_pytorch/calculate_inception_moments.py --dataset C10 --data_root datasets

• Preparing .tgz files for Tensorflow Inception metrics evaluation
• Preparing .npz files for Tensorflow FID metrics evaluation

Operator GANs Training

An example of GAN training command was saved in ./scripts/operator_test.sh. Train a model (cifar10 as an example):

bash ./scripts/operator_test.sh

IE-GAN Training

An example of IE-GAN training command was saved in ./scripts/CIFAR10_32.sh. Train a model (cifar10 as an example):

bash ./scripts/CIFAR10_32.sh

Acknowledgments

• The authors would like to thank Dr. Chaoyue Wang for his assistance.

• Pytorch Inception metrics code from BigGAN-PyTorch.

• TensorFlow Inception Score code from OpenAI's Improved-GAN..

• TensorFlow FID code from TTUR.