This is a Class Project for COMS 4995 Deep Learning at Columbia University, Autumn 2018, by Xu Han, Ruolin Jia, Zhou Zhuang.
Super-Style-Transfer is a toolkit that integrates multiple style transfer models, including CNN based models and GAN based models. It serves as a comprehensive framework to test / run new network structures for style transfer tasks.
Run and tested on Ubuntu 16.04 with python 3.6
Requirements:
python 3.6
pytorch-cpu 0.4.1
torchvision 0.2.1
PIL 5.3.0
Installation:
python setup.py install