//

// This software is released under the MIT License.

// https://opensource.org/licenses/MIT

​

/* ===

ml5 Example

Webcam Image Classification using a pre-trained customized model and p5.js

This example uses p5 preload function to create the classifier

=== */

​

let faceMesh;

let faces = [];

let options = { maxFaces: 1, refineLandmarks: false, flipHorizontal: false };

​

function preload() {

  // Load the faceMesh model

  faceMesh = ml5.faceMesh(options);

}

​

​

// Classifier Variable

let classifier;

// Model URL

let imageModelURL = 'https://teachablemachine.withgoogle.com/models/1JNyrT4b1/' ;

​

// Video

let video;

let flippedVideo;

// To store the classification

let label = "";

​

// Load the model first

function preload() {

  classifier = ml5.imageClassifier(imageModelURL + 'model.json');

​

​

}

​

function setup() {

  createCanvas(320, 260);

  // Create the video

  video = createCapture(VIDEO);

  video.size(320, 240);

  video.hide();

​

  flippedVideo = ml5.flipImage(video)

  // Start classifying

  classifyVideo();

​

  flippedVideo = ml5.flipImage(video)

  classifier.classify(flippedVideo, gotResult);

}

​

​

​

function draw() {

  background(0);

  // Draw the video

  image(flippedVideo, 0, 0);

​

  if(label == "laban up"){

    // Draw the webcam video

  image(video, 0, 0, width, height);

  }

  // Draw all the tracked face points

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

    let face = faces[i];

    for (let j = 0; j < face.keypoints.length; j++) {

      let keypoint = face.keypoints[j];

      fill(0, 255, 0);

      noStroke();

      //texture("💀");

      //textureWrap(0,1,1,1);

      circle(keypoint.x, keypoint.y, 2);

 //   image(winter, 0,0, 320, 240);

 // } else if(label == "alain up"){

  //  image(frankenstein,0,0, 320, 240);

 // } else {

//    image(astronomen, 0, 0, 320, 240);

  }}

​

  // Draw the label

  fill(255);

  textSize(16);

  textAlign(CENTER);

  text(label, width / 2, height - 4);

}

​

 // Get a prediction for the current video frame

function classifyVideo() {

  flippedVideo = ml5.flipImage(video)

  classifier.classify(flippedVideo, gotResult);

}

​

// When we get a result

function gotResult(error, results) {

  // If there is an error

  if (error) {

    console.error(error);

    return;

  }

  // The results are in an array ordered by confidence.

  // console.log(results[0]);

  label = results[0].label;

  // Classifiy again!

  classifyVideo();

}