kqwert11/arbitrary_style_transfer

project-team-vizzun created by GitHub Classroom

Arbitrary Style Transfer

Arbitrary-Style-Per-Model Fast Neural Style Transfer Method

Description

A Pytorch implementation of the 2017 Huang et. al. paper "Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization" https://arxiv.org/abs/1703.06868

This Encoder-AdaIN-Decoder architecture - Deep Convolutional Neural Network as a Style Transfer Network (STN) which can receive two arbitrary images as inputs (one as content, the other one as style) and output a generated image that recombines the content and spatial structure from the former and the style (color, texture) from the latter without re-training the network.

How to run

• Download the Vgg model from here and place it into the models folder.

• Download the pretrained model from here and place it into the models folder.

• Detailed help about running can be found by python3 main.py -h

• For training, the dataset folder structure has to be: content/1/*.jpg and style/1/*.jpg