neural-bauhaus-style-transfer

This repository hosts the code of the students' project Neural Bauhaus Style Transfer. It took place in the winter term 2019/20 and was supervised by Christian Benz (christian.benz@uni-weimar.de).

Final presentation Demo

Results

Contributors

Christian, Xiaoni Cai, Sneha Mohanty, Anh Phuong Le (Annie).

Final deliveries

All codes, datasets and results are saved in project@herder:/media/project/adf39e9c-fcfe-4e8b-bcfc-5c33283d4bce

• Datasets:

1. Processed: Raw data
2. Ready to train: Datasets split into train and test
3. Result train: Result of training the model
4. Result test: Result of testing the model

• Webapp:

1. pytorch-CycleGAN-and-pix2pix-webapp: modified repo for the webapp
2. python-image-flask.zip: python code to run a webapp for trying out the model (see final presentation)
3. To run:
   • put pytorch-CycleGAN-and-pix2pix-webapp and python-image-flask in the same directory
   • cd python-image-flask
   • python app.py

• Webcam: pytorch-CycleGAN-and-pix2pix-webcam for running the webcam demo of some models