Image colorizer using Algorithmia's API

• This project uses Algorithmia to automatically colorize pictures., based on a pre-trained color model.

• Flask was chosen as backend to interact with the local server rendering the picture in your host’s favorite navigator because it is lightweight.

The frontend was build with AngularJS and Bootstrap. Toni Pohl contributed most of that code, as we were both enrolled in the MIRI CS Master program at UPC in Barcelona, Spain.

Slow start:

A step by step guide is included in this repo: see Manual.pdf

Quick start:

Build a Docker image and start a container

Your project directory should contain the Dockerfile for the container to be built. To build the Docker comtainer image just run this in your project directory:

    docker build -t CONTAINER_NAME:CONTAINER_TAG

where "CONTAINER_NAME" and "CONTAINER_TAG" are chosen by user (lowercase only); e.g. colorize_me:dockerfile

To start the container simply enter:

docker run -d -p 8080:5000 -e API_KEY=API_KEY CONTAINER_NAME:CONTAINER_TAG

where 5000 is the flask server (Backend)'s default communication port and 8080 is the chosen port on the container side for the GUI.

Otherwise you could simply use the pre-built k4l4m/colorizer image:

docker run -d -p 8080:5000 -e API_KEY=YOUR_API_KEY k4l4m/colorizer

Short Licensing terms:

You can use the code as you please pursuant to the Licensing terms provided in this repo. In particular you can distribute the content of this repo as long as you distribute the afore-mentioned license with it and you mention its two authors.