Awesome

Creative Machine

A Machine Learning library for Processing. Visit Creative Machine website for more information about the library.

Install

Install using the contribution manager in Processing application.

• Go to Sketch > Import Library > Manage Libraries... and search Creative Machine.

API Reference

Visit our documentation website 🤖

Build

• First, install Adoptium OpenJDK 17 (required by Processing 4+)

Run gradle to build a new release package under /release/creative_machine.zip:

# windows
gradlew.bat releaseProcessingLib

# mac / unix
./gradlew releaseProcessingLib

Developing in IntelliJ IDEA

The library can be imported as an IntelliJ project following the steps below:

• Download and install IntelliJ IDEA
• Clone this repo and build the library following the instructions in the previous section
• Clone the processing4 repo
• Create new project in IntelliJ with the name and location of your choice, for example ml-dev
• Create new module in the project for core Processing, using as content root and the module file location the core folder under the processing4 repo. As "JARs or Directory" dependency, add <path to processing4 repo>/core/library
• Create another module in the project, this time for Creative Machine. Use the Creative Machine's root folder as the content root and module file location. Add the processing-core module as module dependency for this module, and the libs subdirecotry inside the Creative Machine directory (it should have been created during library building step) as itss "JARs or Directory" dependency
• Add the proccessing-core and Creative Machine modules as dependencies in the main module of the project (ml-dev)
• You can now create a test program in under the main module of the project, for example the following code will apply a pre-generated object detection model on an input image:

import processing.core.*;
import ml.*;

public class DetectTest extends PApplet {
    ObjectDetector detector;
    PImage img;

    public void settings() {
        size(parseInt(args[0]), parseInt(args[1]));
    }

    public void setup() {
        detector = new ObjectDetector(this, "coco_ssd");
        img = loadImage("dog_bike_car.jpeg");
        MLObject[] output = detector.detect(img, "output.png");
        // print a label and confidence score of each object
        for (int i = 0; i < output.length; i++) {
            println(output[i].getLabel() + " detected! (confidence: " + output[i].getConfidence() + ")");
        }
    }

    public void draw() {
       // draw a bounding box of each object
       image(img, 0, 0);
       noFill();
       stroke(255, 0, 0);
       for (int i = 0; i < output.length; i++) {
           MLObject obj = output[i];
           rect(obj.getX(), obj.getY(), obj.getWidth(), obj.getHeight());
       }
    }

    static public void main(String[] args) {
      PApplet.main(DetectTest.class, "768", "576");
    }
}

Please note that the input image should be placed inside the subdirectory data located inside the root of the IntelliJ project (i.e.: ml-dev/data)