ARRPNGAN

Pytorch implementation for reproducing ARRPNGAN results in the paper "ARRPNGAN: Text-to-Image GAN with Attention Regularization and Region Proposal Networks" by Fengnan Quan, Bo Lang and Yanxi Liu.

Installation

• Install PyTorch and dependencies from http://pytorch.org
• Install Torch vision from the source.

Requirements

• python: 3.7
• pytorch>=0.4.1
• install pycocotools:

pip install "git+https://github.com/cocodataset/cocoapi.git#subdirectory=PythonAPI"

-others:

pip install -r requirements.txt

Train/Test

• Clone this repo:

git clone https://github.com/QuanFN/ARRPNGAN.git
cd ARRPNGAN

• Download our preprocessed data from here and save them to data/
• Download the birds image data. Extract them to data/birds/
• Download coco dataset and extract the images to data/coco/
• Train a model on CUB:

./train_bird.sh

• Test a model on CUB:

./test_bird.sh

• Train a model on COCO:

./train_coco.sh

• Test a model on COCO:

./test_coco.sh

Examples generated by ARRPNGAN

Citation

If you use this code for your research, please cite our paper.