Project Information:
Author: Ivana Moreno (me), Emilia Kaitazoff
Teacher: Matteo Loglio, Danilo di Cuia
Course: CAS 2 - Machine Learning for Creatives
SUPSI, MaInD, 2019/20
Tools used:
Vector diagrams: Adobe Illustrator
Image editing: Adobe Photoshop
Prototype: Machine Learning -
Runwayml, Keras OCR and p5.js, HTML, CSS,
JavaScript, monitor, webcam and printed color signs
Runwayml, Keras OCR and p5.js, HTML, CSS,
JavaScript, monitor, webcam and printed color signs
Video editing: Adobe Premiere
Abstract
A brush is a Brush consists of an art expression tool that works in collaboration with a machine. Through optical character recognition (OCR) is capable of turning different words like brush, magenta, cyan, green and red (among others) into a visual outcome, generating a digital painting on the computer that’s running the code.
Interface and interaction modality
The user will stand in front of a camera connected to a computer and will interact with it by showing tangible objects, in this case a brush and printed labels with colors' names.
It will take as input the different words, and through Keras OCR model the machine will recognize the text written on the labels projecting them on a screen as the user moves the brush.
User experience
The user will play with the tool through intuitive hand movements and by showing the label with chosen color's name will switch the brush’s color.
Research and development context
In a future development the machine will be able to recognize a broader range of words in order to provide more properties for the brush.
Technology
The technology used to generate the digital painting were Runwayml, Keras OCR model and p5.js.
