Deep-Color

Colorization of Images using CNN and GAN.

• Go through the Dependencies folder to install whatever is needed

• to_Gray python script can be used to convert a color image to Grayscale

• Following are the Dependencies to be installed:

Instructions

Setting up Dependencies

Installing OpenCV

• Run the following commands

  pip install opencv-contrib-python
  

Custom installting all dependencies for GPU

Installing Cuda-9.0

• Download the cuda runfile here

• Install some other dependencies and then install cuda

  sudo apt-get install freeglut3-dev build-essential libx11-dev libxmu-dev libxi-dev libgl1-mesa-glx libglu1-mesa libglu1-mesa-dev
  sudo chmod 777 *.runfile  # '*.runfile' denotes the file name you just downloaded
  sudo ./cuda_9.0.176_384.81_linux.run -toolkit -samples -silent -override #
  

• Create a symbolic link to cuda to avoid missing library errors

  cd /usr/local
  sudo ln -s /usr/local/cuda-9.0 cuda
  

• Lower the gcc version of the system to before 6

  gcc --version # check the gcc version
  sudo apt install gcc-5 g++-5
  sudo update-alternatives --install /usr/bin/gcc gcc /usr/bin/gcc-5 50
  sudo update-alternatives --install /usr/bin/g++ g++ /usr/bin/g++-5 50
  

• Set environment variables

   export PATH=/usr/local/cuda-9.0/bin${PATH:+:${PATH}}
   export LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:/usr/local/cuda-9.0/lib64
  

• Also modify and add path to .bashrc just in case

  export LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:/usr/local/cuda-9.0/lib64
  

• Verify cuda-9.0 installation

  cd NVIDIA_CUDA-9.0_Samples/5_Simulations/fluidsGL
  make clean && make
  ./fluidsGL
  

  If cuda-9.0 has been installed properly, there should be no error messages during making. After running you will the fluid window.

Installing cuDNN

• Go to this page and create an account

• Click “Download cuDNN v7.0.5 (Dec 5, 2017), for CUDA 9.0” and download the following files: runtime library, developer library, and code samples and user guide.

• Run the following commands to install

  sudo dpkg -i libcudnn7_7.0.5.15-1+cuda9.0_amd64.deb
  sudo dpkg -i libcudnn7-dev_7.0.5.15-1+cuda9.0_amd64.deb
  sudo dpkg -i libcudnn7-doc_7.0.5.15-1+cuda9.0_amd64.deb
  

• Install Freeimage and verify cuDNN

  sudo apt-get install libfreeimage3 libfreeimage-dev
  cp -r /usr/src/cudnn_samples_v7/ $HOME
  cd $HOME/cudnn_samples_v7/mnistCUDNN
  make clean && make
  ./mnistCUDNN
  

Installing Tensorflow

• We'll need to setup a virtual environment and then install.

  sudo apt-get install libcupti-dev
  export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda/extras/CUPTI/lib64
  
  sudo apt-get install python3-pip python3-dev python-virtualenv
  virtualenv --system-site-packages -p python3 tensorflow # create a enviroment named tensorflow
  

• Installing TensorFlow CPU

  source ~/tensorflow/bin/activate
  pip3 install --upgrade tensorflow  # install the cpu version
  

• Verifying TensorFlow CPU Make sure you are in the same TF environment. Enter python

  import tensorflow as tf
  hello = tf.constant('Hello, TensorFlow!')
  sess = tf.Session()
  print(sess.run(hello))
  

• Installing TensorFlow GPU Make sure you are in the same environment

  pip3 install --upgrade tensorflow-gpu
  

  Verify as before

Install PyTorch

• First we need to install Anaconda. Click here to download the file.

  bash ~/Downloads/'file which was just downloaded'
  

  Agree to the TnC. Let it set Path automatically. Wait for installations to finish. Install VSCode if you want, else you can skip. Run the following:

  source ~/.bashrc
  

  In another teminal run:

  anaconda-navigator
  

  If Navigator opens up, Anaconda has been installed successfully.

• Do the following steps Check your group by typing this

  groups
  

  The first group is usally the group right now in use Also keep your username in mind. And now execute the following command:

  chown -R YOUR_group:YOUR_USER_name anaconda3
  

• Set channels, and download pytorch from mirror link (faster)

  conda config --add channels https://mirrors.ustc.edu.cn/anaconda/pkgs/free/
  conda config --add channels https://mirrors.ustc.edu.cn/anaconda/pkgs/main/
  conda config --set show_channel_urls yes
  conda install pytorch torchvision cuda91 -c https://mirrors.ustc.edu.cn/anaconda/cloud/pytorch/
  

  Open a python script and try to import torch. If it imports, you've installed PyTorch

Installing other dependencies

• Installing Scikit,Numpy,Pandas and other libraries. Make sure pip has been installed in the system.

  pip3 install numpy
  pip3 install pandas
  pip3 install scipy
  pip3 install scikit-learn
  pip3 install matplotlib
  

Install Keras with backend as TensorFlow

• Make sure Tensorflow environment is activate

  pip install -q keras
  pip install http://download.pytorch.org/whl/cu90/torch-0.4.0-cp36-cp36m-linux_x86_64.whl
  pip install torchvision
  sudo apt-get -qq install -y graphviz
  pip install -q pydot
  pip install mxnet-cu90
  pip install scikit-image
  

• Verify installation, by running them in Ipython or some Python script and check version.

  ipython
  import tensorflow as tf
  import torch
  import cv2
  import mxnet
  import keras
  

Installing Anaconda and running Project

• Click here to download the file.

  bash ~/Downloads/'file which was just downloaded'
  

  Agree to the TnC. Let it set Path automatically. Wait for installations to finish. Install VSCode if you want, else you can skip. Run the following:

  source ~/.bashrc
  

  In another teminal run:

  anaconda-navigator
  

  If Navigator opens up, Anaconda has been installed successfully.

• Do the following steps Check your group by typing this

  groups
  

  The first group is usally the group right now in use Also keep your username in mind. And now execute the following command:

  chown -R YOUR_group:YOUR_USER_name anaconda3
  

• Set channels, and download pytorch from mirror link (faster)

  conda config --add channels https://mirrors.ustc.edu.cn/anaconda/pkgs/free/
  conda config --add channels https://mirrors.ustc.edu.cn/anaconda/pkgs/main/
  conda config --set show_channel_urls yes
  conda install pytorch torchvision cuda91 -c https://mirrors.ustc.edu.cn/anaconda/cloud/pytorch/
  

  Open a python script and try to import torch. If it imports, you've installed PyTorch

Installing Tensorflow, Keras and Scikit-Image

conda install -c anaconda keras 
conda install -c anaconda scikit-image