DILTER

Live Movement-Tracking | Public Installation | Image Pixelation

Individual Work

Timeline: November 2019 - December 2019 (3 weeks)

Made with: Software - PoseNet, P5.js, ML5.js, Adobe Illustrator.

| Opportunity

We were asked to develop a user interface on the web or mobile platform utilizing any machine learning models we covered during the class.

| Outcome

Delivered a machine-learning-driven interface that introduces the fundamental concepts of this technology to all-aged-groups, especially children.

DILTER is a movement-based interactive installation combined with generative art through P5.js. This interface applies Disney characters’ features to the user’s bodies. It experiments with live movement-tracking with the use of the ml5.js model PoseNet. With image pixel manipulation, it immerses users in a world of black-and-white animation. Through applying iconic Disney characters’ features, this project hopes to approach all-aged groups, especially children, with the fundamental concepts of machine learning technology in everyday life.

View working documentation here.

This was my first attempt at building an experience with live movement-tracking technology.

This successful experiment led me to continue further research on this topic for my capstone project.

back to Interactive Media