Shakalaka Boom Boom is an A.I. experiment which creates a magical experience by converting anything drawn on a mobile canvas into a real object in augmented reality world. To get better picture about the project, check out the video below
This repository contains the code for the Shakalaka Boom Boom experiment server which has the capability to classify drawing. Currently the machine learning model deployed in the server is capable of classifying drawings of candle, chair, cup, lamp and vase only.
This project makes use of conda to manage dependencies. So make sure you have conda installed before proceeding or install the dependencies manually by looking at the environment.yml file. Once you have conda installed get the dependencies installed by running
conda env create -f environment.yml
- Activate the conda environment by running
source activate shakalakaBoomBoom
(Windows users omit the word source while activating the environment)
- Change into repository directory and then export the
FLASK_APPenvironment variable by running
export FLASK_APP=app.py
- Start local server by running
flask run
Once the local server is running, you will get a url such as http://127.0.0.1:5000/
Then send out POST requests to upload endpoint (for example http://127.0.0.1:5000/upload) to classify the drawing and get the classification result as response.
To expose the server to external world, use a tool such as ngrok. Install it and then run
./ngrok http 5000