this project uses deep learning to bring old tunisian images from history back to life using caffe framework and deploying it with python flask app
**this project is largely dependent on richzhang paper and its pre-trained model**
this project uses deep learning to bring old tunisian images back to life. Making use of an already pretrained model I ceated a flask app that allows users to colorize their own black and white photos as well as explore our gallery.Then I managed to deploy this project on HEROKU ( link: imgcol.herokuapp.com )
**challenges:**
-one of the challenges that I went through during this project is dealing with output images from the model and then using them in the front this issue is caused by image caching in the browser.
-Also since heroku is a saas I had to use custom buildpacks in order to add necessary dependcies to heroku's OS in order for OPENCV to work.
HOSTED PROJECT LINK:
imgcol.herokuapp.com
used links:
*Black and white image colorization with OpenCV and Deep Learning:
This article was very useful to inderstand the research paper and to make use of the model loading it with OPENCV
pyimagesearch.com
@inproceedings{zhang2016colorful,
title={Colorful Image Colorization},
author={Zhang, Richard and Isola, Phillip and Efros, Alexei A},
booktitle={ECCV},
year={2016}
}