let classifier;

let label = '';

​

let x = 200;

let speedX = 0;

​

function modelReady() {

  console.log('Model loaded');

  // note how classify() with soundClassifiers

  // doesn't need the first parameter, instead

  // it automatically uses the microphone as

  // input

  classifier.classify(gotResults);

}

​

function gotResults(error, results) {

  if (!error) {

    //console.log(results[0]);

​

    label = results[0].label + ' @ ' + nf(results[0].confidence * 100, 0, 2) + '%';

​

    if (results[0].label == 'Right' && results[0].confidence > 0.8) {

      speedX = 0.2;

    }

    if (results[0].label == 'Left' && results[0].confidence > 0.8) {

      speedX = -0.2;

    }

  }

  classifier.classify(gotResults);  // run again (loop)

}

​

function setup() {

  createCanvas(400, 400);

​

  // the first parameter is the URL to the model on Teachable Machine

  classifier = ml5.soundClassifier('https://teachablemachine.withgoogle.com/models/_fuC7qYaj/model.json', modelReady);

}

​

function draw() {

  background(255);

​

  x = x + speedX;

  noStroke();

  fill(255, 0, 0, 50);

  ellipse(x, 200, 50, 50);

  fill(0);

  textAlign(CENTER);

  text(label, width/2, height/2);

}

​