Skip to content

Latest commit

 

History

History
15 lines (9 loc) · 1.2 KB

README.md

File metadata and controls

15 lines (9 loc) · 1.2 KB

Dare Mighty Things Hackathon & Conference 2019

JamDraw - using computer vision, machine learning, and data science


Correct file for drive is named as FinalProjDriver.py ' Visual Demonstration coming soon '

We created a web application that takes your stories (drawings) and generates a song based off the percentage of a dominant color in your drawing.

To create this, we started off with rapid prototyping harnessing the design thinking skills. This lead us to come up with such a unique idea: JamDraw.

Our web application is run through a flask front end and uses OpenCV (Computer Vision Library) to detect and track an object. It uses Pandas to filter through spotify song data to find the attribute "danceability" for a song.

We also incorporated machine learning using an unsupervised learning method: KMeans Clustering. KMeans Clustering was used to determine the dominant color in the image by narrowing down to three clusters and creating a histogram of the ratio of the three most dominant colors in the image. A percentage was associated with the second color out of the three which then translated to a value on the danceability scale which is an attribute of a song through the Spotify API.