let poseNet;

let poses = [];

​

let speechRecognition = new webkitSpeechRecognition();

let words = [];

​

function setup() {

  createCanvas(640, 480);

  video = createCapture(VIDEO);

  video.size(width, height);

​

  poseNet = ml5.poseNet(video, modelReady);

  poseNet.on('pose', function(results) {

    poses = results;

  });

  video.hide();

​

  speechRecognition.continuous = true;

  speechRecognition.interimResults = true;

  speechRecognition.onresult = gotSpeech;

  speechRecognition.start();

}

​

function draw() {

  image(video, 0, 0, width, height);

​

  if (poses.length > 0) {

    let pose = poses[0].pose;

    let keypoint = pose.keypoints[0]; // nose

    if (keypoint.score > 0.2) {

      fill(255, 0, 0);

      noStroke();

      ellipse(keypoint.position.x, keypoint.position.y, 10, 10);

​

      for (let i = 0; i < words.length; i++) {

        fill(255);

        textSize(random(10, 24));

        text(words[i], keypoint.position.x + random(-100, 100), keypoint.position.y + random(-100, 100));

      }

    }

  }

}

​

function gotSpeech(event) {

  let latest = event.results[event.results.length - 1];

  if (latest.isFinal) {

    words = latest[0].transcript.split(' ');

  }

}

​

function modelReady() {

  select('#status').html('Model Loaded');

}

​