I created a CNN model that converts black-and-white images to colored images using Python and Tensorflow. This project was implemented by studying Emill Wallner's "Coloring-greyscale-images" git repository. Here is the link for the said repository -> https://github.com/emilwallner/Coloring-greyscale-images
PLEASE READ THIS BELOW
- The testing and training images of the 'dataset' directory were directly taken from Emil's git repository, you may do the same.
- The 'dataset800' directory images were taken from 2014mchidamb's DeepColorization repository. The link to the dataset is provided below. https://github.com/2014mchidamb/DeepColorization/tree/master/face_images
- To download the above dataset, I used TortoiseSVN, the tutorials for which, you will find on youtube.
- If you are to implement this project by training the model yourself, Google Collabs wont be sufficient. You will have to train the model on your own PC using Anaconda and Jupyter notebook.
- Please note that if you are going to use Jupyter notebook, you will have to download all the libraries needed into your PC
- You can do this by opening Anaconda Prompt, and excuting pip install commands.
- You can find out the individual pip install commands for all libraries needed on google.
8.If you wish to just use the model i trained and implement your project, please be patient as I have not yet uploaded the .h5 files you will need.
THAT BEING SAID, ALL THE BEST!!!! I HOPE YOU HAVE A WONDERFUL TIME IMPLEMENTING THIS PROJECT.
GOOD LUCK!