Code and data for paper "High-Resolution Network for Photorealistic Style Transfer".
This code is based on PyTorch 1.0.0 and Cuda 9.0.
Here are some results from our algorithm (from left to right are input, style and our output):
Our work is inspired by Deep High-Resolution Representation Learning for Human Pose Estimation.
The transfer code is based on Udacity.
Feel free to contact me if there is any question (Ming Li limingcv@gmail.com).