// the json file (model topology) has a reference to the bin file (model weights)

const checkpointURL = modelURL + "model.json";

// the metatadata json file contains the text labels of your model and additional information

const metadataURL = modelURL + "metadata.json";

​

​

const flip = true; // whether to flip the webcam

​

let model;

let totalClasses;

let myCanvas;

​

let classification = "None Yet";

let probability = "100";

let poser;

let video;

​

let clapimage;

let dabimage;

let lassoimage;

let handsimage;

​

let dancer;

let dummy;

​

let yay;

let aw;

​

let score = "0"

let good = (score + 10);

​

var songs = ["one, two, three"];

​

function preload(){

  songs[0] = loadSound('American Boy Dance.mp3');

  clapimage = loadImage("Clapping Icon.png");

  dabimage = loadImage("Dab Icon.png");

  lassoimage = loadImage("Lasso Icon.png");

  handsimage = loadImage("Hands Up Icon.png");

  dancer = loadImage("Nice Dancing.png");

  dummy = loadImage("Practicem.png");

  yay = loadImage("Great.png");

  aw = loadImage("Bad.png");

}

// A function that loads the model from the checkpoint

async function load() {

  model = await tmPose.load(checkpointURL, metadataURL);

  totalClasses = model.getTotalClasses();

  console.log("Number of classes, ", totalClasses);

}

​

​

async function setup() {

  myCanvas = createCanvas(400, 400);

  // Call the load function, wait until it finishes loading

  videoCanvas = createCanvas(320, 240)

​

  await load();

  video = createCapture(VIDEO, videoReady);

  video.size(320, 240);

  video.hide();

​

}

​

function draw() {

  background(255);

  if(video) image(video,0,0);

  //rect(0, 200, 200, 50)

  textSize(16);

  //text("Result:" + classification, 10, height -20);

    if (songs[0].isPlaying() == false){

  songs[0].play();

    }

  if(millis()== 10){

    fill(255, 0, 0);

    circle(100, 100, 100, 100);

  }

  text("score", 20, 20)

  text(score, 70, 20)

​

  //text("Probability:" + probability, width/2, height -20)

  ///ALEX insert if statement here testing classification against apppropriate part of array for this time in your video

  if((millis() >= 8000)&&(millis() <= 12000)){

    //rect(100, 100, 100, 50);

    image(handsimage, 200, 130, 100, 100);

    correctornot();

  }

  if((millis() >= 12000)&&(millis() <= 16000)){

        //rect(100, 100, 100, 50);

    image(dabimage, 200, 130, 120, 100);

    correctornot2();

  }

  if((millis() >= 16000)&&(millis() <= 20000)){

        //rect(100, 100, 100, 50);

    image(lassoimage, 200, 130, 120, 100);

    correctornot3();

  }  

  if((millis() >= 20000)&&(millis() <= 24000)){

        //rect(100, 100, 100, 50);

    image(handsimage, 200, 130, 100, 100);

    correctornot();

  }

    if((millis() >= 24000)&&(millis() <= 28000)){

        //rect(100, 100, 100, 50);

    image(clapimage, 200, 130, 100, 100);

    correctornot5();

    }

    if((millis() >= 28000)&&(millis() <= 32000)){

        //rect(100, 100, 100, 50);

    image(handsimage, 200, 130, 100, 100);

    correctornot();

  }

    if((millis() >= 32000)&&(millis() <= 36000)){

        //rect(100, 100, 100, 50);

    image(lassoimage, 200, 130, 120, 100);

    correctornot3();

  }

    if((millis() >= 36000)&&(millis() <= 40000)){

        //rect(100, 100, 100, 50);

    image(dabimage, 200, 130, 120, 100);

    correctornot2();

  }

      if((millis() >= 41000)&&(millis() <= 80000)){

goodorbad();

  }

​

  textSize(8);

  if (poser) { //did we get a skeleton yet;

    for (var i = 0; i < poser.length; i++) {

      let x = poser[i].position.x;

      let y = poser[i].position.y;

      //ellipse(x, y, 5, 5);

      //text(poser[i].part, x + 4, y);

    }

  }

​

}

​

function goodorbad(){

  textAlign(CENTER);

  textSize(20);

  fill(255);

  text("Your Score :" + score, width/2, 50);

  //text(score, width/2, 50);

  if(score >= 450){

    fill(255);

    stroke(255);

    rect(75, 75, 200, 100);

    image(dancer, 100, 90, 150, 75);

  }

  else{

    fill(255);

    stroke(255);

    rect(70, 75, 200, 100);

    image(dummy, 100, 90, 150, 75);

  }

}

function correctornot(){

  if(classification == "Handsup"){

    image(yay, 210, 80, 75, 50);

    print("good!");

    score++

  }else{

    image(aw, 210, 80, 75, 50);

    print("bad");

  }

}

function correctornot2(){

  if(classification == "Dab"){

    image(yay, 210, 80, 75, 50);

    print("good!");

    score++

  }else{

    image(aw, 210, 80, 75, 50);

    print("bad");

    score--

  }

}

function correctornot3(){

  if(classification == "Cowboy"){

    image(yay, 210, 80, 75, 50);

    print("good!");

    score++

  }else{

    image(aw, 210, 80, 75, 50);

    print("bad");

    score--

  }

}

function correctornot4(){

  if(classification == "Knees Touch"){

    image(yay, 210, 80, 75, 50);

    print("good!");

    score++

  }else{

    image(aw, 210, 80, 75, 50);

    print("bad");

    score--

  }

}

function correctornot5(){

  if(classification == "Clapping"){

    image(yay, 210, 80, 75, 50);

    print("good!");

    score++

  }else{

    image(aw, 210, 80, 75, 50);

    print("bad");

    score--

  }

}

​

​

function videoReady() {

  console.log("Video Ready");

  predict();

}

​

​

async function predict() {

  // Prediction #1: run input through posenet

  // predict can take in an image, video or canvas html element

  const flipHorizontal = false;

  const {

    pose,

    posenetOutput

  } = await model.estimatePose(

    video.elt, //webcam.canvas,

    flipHorizontal

  );

  // Prediction 2: run input through teachable machine assification model

  const prediction = await model.predict(

    posenetOutput,

    flipHorizontal,

    totalClasses

  );

​

  // console.log(prediction);

  // Sort prediction array by probability

  // So the first classname will have the highest probability

  const sortedPrediction = prediction.sort((a, b) => -a.probability + b.probability);

​

  //communicate these values back to draw function with global variables

  classification = sortedPrediction[0].className;

  probability = sortedPrediction[0].probability.toFixed(2);

  if (pose) poser = pose.keypoints; // is there a skeleton

  predict();

}