Skip to content
master
Go to file
Code

Latest commit

 

Git stats

Files

Permalink
Failed to load latest commit information.
Type
Name
Latest commit message
Commit time
 
 
 
 
 
 
 
 
 
 

README.md

IMA_RWCNN_project

Project on the texture synthesis using random filters in a CNN paper in the digital imaging course (MVA 2018-2019).

WORK IN PROGRESS

Our goal is to reproduce the method from the texture synthesis using random filters in a CNN paper using the code from Gatys as described in the paper "Texture Synthesis Using Convolutional Neural Networks" (Gatys et al., NIPS 2015) http://arxiv.org/abs/1505.07376.

Contents

  • reviews
  • Final_Notebook.ipynb contains the code used to produce the report results. It is a notebook meant to be run in Google Colaboratory, hence it might need a few tweaks to run properly on a local instance.

Requirements

This repository is developed on Linux 64-bit in Python 3.7 using (Mini)conda.

To instanciate the conda environment, run conda create --file conda-env.txt --name ima.

Then use pip to install the required Python packages: pip -r requirements.txt.

To use the virtual environment: source activate ima

To install a custom kernel in jupyter lab/notebook: ipython kernel install --user --name=ima

To work on Jupyter notebook or Jupyter lab you will need to install it via pip. Follow the instructions here to make the environment kernel available in your jupyter installation.

Old

Docker instructions to run Gatys code

See detailed issue

  1. Install Docker: https://docs.docker.com/engine/installation/
  2. sudo docker pull leongatys/caffe-cpu (download the image)
  3. Download the code (git clone https://github.com/leongatys/DeepTextures.git)
  4. Download the fully trained VGG network: http://bethgelab.org/media/uploads/deeptextures/vgg_normalised.caffemodel
  5. Start a notebook server in the docker container: sudo docker run -v $HOME:/home -p 8888:8888 -d leongatys/caffe-cpu jupyter-notebook
  6. Check the docker container IP: sudo docker inspect -f '{{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}' <docker_id> where docker_id is given at the end of the previous command or by sudo docker container ls -aq
  7. Go to the browser and type in the IP with port 8888 (<output_IP>:8888)
  8. You can play with the existing notebook and/or create your own.
  9. To stop the docker container: sudo docker container stop <container_hash>

About

Project on the texture synthesis using random filters in a CNN paper in the digital imaging course

Resources

License

Releases

No releases published

Packages

No packages published
You can’t perform that action at this time.