Skip to content

Latest commit

 

History

History
6 lines (6 loc) · 624 Bytes

README.md

File metadata and controls

6 lines (6 loc) · 624 Bytes

Attention-GAN

This repository provides the PyTorch code for our paper “Attention-GAN for object transfiguration in wild images”(ECCV2018). This code is based on the PyTorch (0.4.1) implementation of CycleGAN. You may need to train several times as the quality of the results are sensitive to the initialization.

Data Preparation

bash datasets/download_cyclegan_dataset.sh horse2zebra

Train

bash scripts/train_attngan.sh