Spaces:

YOUSEF2434
/

Computer-Vision-Lab

Paused

File size: 14,995 Bytes

a566fb0

<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>Enhanced Object Detection</title>
<link href="https://unpkg.com/material-components-web@latest/dist/material-components-web.min.css" rel="stylesheet">
<style>

  body { font-family: 'Roboto', sans-serif; margin: 2em; color: #3d3d3d; background: #f0f2f5; }

  h1 { color: #007f8b; text-align: center; }

  .container { max-width: 1200px; margin: 0 auto; background: white; padding: 20px; border-radius: 12px; box-shadow: 0 4px 6px rgba(0,0,0,0.1); }

  

  /* Controls Section */

  .controls { display: flex; gap: 20px; flex-wrap: wrap; margin-bottom: 20px; padding: 15px; background: #e6fcfd; border-radius: 8px; align-items: center; }

  .control-group { display: flex; flex-direction: column; min-width: 200px; }

  label { font-weight: bold; font-size: 0.9em; margin-bottom: 5px; color: #007f8b; }

  input[type=range] { width: 100%; }

  .upload-btn { background: #007f8b; color: white; padding: 10px 20px; border-radius: 25px; cursor: pointer; display: inline-block; font-weight: bold; text-align: center; }

  .upload-btn:hover { background: #006069; }

  input[type="file"] { display: none; }



  /* Image Grid */

/* Replace the existing .image-grid and .detect-card styles */



.image-grid {

  /* Disable Grid, use Columns instead */

  display: block; 

  column-count: 3; /* Creates 3 columns like Pinterest */

  column-gap: 20px;

}



/* Responsive: 2 columns on smaller screens */

@media (max-width: 900px) {

  .image-grid { column-count: 2; }

}

@media (max-width: 600px) {

  .image-grid { column-count: 1; }

}



.detect-card { 

  position: relative; 

  border-radius: 8px; 

  overflow: hidden; 

  box-shadow: 0 2px 4px rgba(0,0,0,0.1); 

  background: #000; 

  cursor: pointer; 

  transition: transform 0.2s; 

  z-index: 1; 

  

  /* NEW: Prevents card from splitting across columns */

  break-inside: avoid-column; 

  margin-bottom: 20px; 

}



  .detect-card:hover { transform: scale(1.01); }

  .detect-card img { display: block; width: 100%; height: auto; transition: opacity 0.3s; }

  

  /* Processing State */

  .detect-card.processing { pointer-events: none; } /* Prevent double clicks */

  .detect-card.processing img { opacity: 0.6; filter: grayscale(50%); }



  /* --- NEW: Inference Loading Bar --- */

  .inference-panel {

    position: absolute;

    bottom: 0;

    left: 0;

    width: 100%;

    background: rgba(255, 255, 255, 0.95);

    padding: 15px;

    box-sizing: border-box;

    transform: translateY(100%); /* Hidden by default */

    transition: transform 0.3s cubic-bezier(0.4, 0.0, 0.2, 1);

    z-index: 50;

    border-top: 3px solid #007f8b;

  }

  

  /* Show panel when processing */

  .detect-card.processing .inference-panel {

    transform: translateY(0);

  }



  .inference-status {

    display: flex;

    justify-content: space-between;

    font-weight: bold;

    color: #007f8b;

    margin-bottom: 8px;

    font-size: 0.9rem;

  }



  .progress-track {

    width: 100%;

    height: 6px;

    background: #e0e0e0;

    border-radius: 3px;

    overflow: hidden;

  }



  .progress-bar {

    height: 100%;

    background: #007f8b;

    width: 30%;

    border-radius: 3px;

    animation: loading 1.5s infinite ease-in-out;

  }



  @keyframes loading {

    0% { transform: translateX(-100%); }

    100% { transform: translateX(400%); }

  }

  /* ---------------------------------- */

  

  /* Bounding Boxes */

  .highlighter { position: absolute; border: 2px solid; border-radius: 4px; z-index: 10; pointer-events: none; }

  .label-tag { position: absolute; padding: 2px 6px; color: white; font-size: 11px; font-weight: bold; border-radius: 4px; pointer-events: none; z-index: 11; white-space: nowrap; box-shadow: 0 1px 2px rgba(0,0,0,0.2); }

  

  /* Loading Spinner (Initial Model Load) */

  #loader { position: fixed; top: 50%; left: 50%; transform: translate(-50%, -50%); padding: 20px; background: white; border-radius: 8px; box-shadow: 0 0 20px rgba(0,0,0,0.2); display: none; z-index: 1000; font-weight: bold; }

</style>
</head>
<body>

<div class="container">
  <h1>Smart Object Recognition</h1>
  
  <div class="controls">
    <div class="control-group">
      <label for="imageUpload" class="upload-btn">📂 Upload Image</label>
      <input type="file" id="imageUpload" accept="image/*">
    </div>
    
    <div class="control-group">
      <label>Confidence Threshold: <span id="confValue">50</span>%</label>
      <input type="range" id="confidenceSlider" min="10" max="90" value="50">
      <small>Increase to remove weak guesses.</small>
    </div>

    <div class="control-group">
      <label>Overlap Fix (NMS): <span id="overlapValue">30</span>%</label>
      <input type="range" id="overlapSlider" min="0" max="100" value="30">
      <small>Lower value = Fewer overlapping boxes.</small>
    </div>
  </div>

  <div id="loader">Loading AI Model...</div>

  <div class="image-grid" id="imageContainer">
    <!-- Image 1 -->
    <div class="detect-card">
      <img src="https://assets.codepen.io/9177687/coupledog.jpeg" crossorigin="anonymous" />
    </div>
    
    <!-- Image 2 -->
    <div class="detect-card">
      <img src="https://assets.codepen.io/9177687/doggo.jpeg" crossorigin="anonymous" />
    </div>

    <!-- Image 3 (FIXED: Added wrapper div) -->
    <div class="detect-card">
      <img src="https://tse3.mm.bing.net/th/id/OIP.mIJJ36cXpVujF1wnZnd4VQHaE8?rs=1&pid=ImgDetMain&o=7&rm=3" crossorigin="anonymous" />
    </div>

    <!-- Image 4 -->
    <div class="detect-card">
      <img src="https://images.pexels.com/photos/23409055/pexels-photo-23409055/free-photo-of-cars-on-street-in-town.jpeg?auto=compress&cs=tinysrgb&w=1260&h=750&dpr=1" crossorigin="anonymous" />
    </div>

    <!-- Image 5 -->
    <div class="detect-card">
      <img src="https://tse4.mm.bing.net/th/id/OIP.bWwaHeR-aoBb3esBRaAEEgHaE8?rs=1&pid=ImgDetMain&o=7&rm=3" crossorigin="anonymous" />
    </div>

    <!-- Image 6 -->
    <div class="detect-card">
      <img src="https://tse4.mm.bing.net/th/id/OIP.vs_d1C-7n4PoNv0GVlaVDwHaFj?rs=1&pid=ImgDetMain&o=7&rm=3" crossorigin="anonymous" />
    </div>

    <!-- Image 7 -->
    <div class="detect-card">
      <img src="https://tse4.mm.bing.net/th/id/OIP.V1zVa5IUI22o0i6gG4or2QHaLH?rs=1&pid=ImgDetMain&o=7&rm=3" crossorigin="anonymous" />
    </div>

    <!-- Image 8 -->
    <div class="detect-card">
      <img src="https://th.bing.com/th/id/R.2be55af1ab4a38df1a9b54bf6b68a8bd?rik=f7wiKdc8N42mgA&pid=ImgRaw&r=0" crossorigin="anonymous" />
    </div>

    <!-- Image 9 -->
    <div class="detect-card">
      <img src="https://images.pexels.com/photos/20625972/pexels-photo-20625972.jpeg?cs=srgb&dl=pexels-saturnus99-20625972.jpg&fm=jpg" crossorigin="anonymous" />
    </div>

    <!-- Image 10 -->
    <div class="detect-card">
      <img src="https://tse4.mm.bing.net/th/id/OIP.ZMnNqw1GVTa9HpHvsTYcjQAAAA?rs=1&pid=ImgDetMain&o=7&rm=3" crossorigin="anonymous" />
    </div>

    <!-- Image 11 -->
    <div class="detect-card">
      <img src="https://bestbackpacklab.com/wp-content/uploads/2021/05/children-1536x864.jpg" crossorigin="anonymous" />
    </div>

    <!-- Image 12 -->
    <div class="detect-card">
      <img src="https://tse1.explicit.bing.net/th/id/OIP.za2l0WGKXbR4Qkj8phu2UwHaE8?rs=1&pid=ImgDetMain&o=7&rm=3" crossorigin="anonymous" />
    </div>

    <!-- Image 13 -->
    <div class="detect-card">
      <img src="https://tse2.mm.bing.net/th/id/OIP.bcOP7ZTpLAyyl3tKpdk5gAHaFB?rs=1&pid=ImgDetMain&o=7&rm=3" crossorigin="anonymous" />
    </div>

    <!-- Image 14 -->
    <div class="detect-card">
      <img src="https://tse3.mm.bing.net/th/id/OIP.PC6Fr2mEuGUsEiaNfCSOaAHaE7?rs=1&pid=ImgDetMain&o=7&rm=3" crossorigin="anonymous" />
    </div>

    <!-- Image 15 -->
    <div class="detect-card">
      <img src="https://tse3.mm.bing.net/th/id/OIF.EABwKojMHBX0uEfpxor95w?rs=1&pid=ImgDetMain&o=7&rm=3" crossorigin="anonymous" />
    </div>

    <!-- Image 16 -->
    <div class="detect-card">
      <img src="https://tse4.mm.bing.net/th/id/OIP.qliYrfiREN-ydW4DxWYfSgHaE7?w=626&h=417&rs=1&pid=ImgDetMain&o=7&rm=3" crossorigin="anonymous" />
    </div>
  </div>
</div>

<script type="module">

import { ObjectDetector, FilesetResolver } from "https://cdn.jsdelivr.net/npm/@mediapipe/tasks-vision@0.10.2";



const loader = document.getElementById("loader");

let objectDetector;

let runningMode = "IMAGE";



// SETTINGS

let confidenceThreshold = 0.5;

let overlapThreshold = 0.3; 



// 1. Initialize MediaPipe

const initializeObjectDetector = async () => {

  loader.style.display = "block";

  const vision = await FilesetResolver.forVisionTasks("https://cdn.jsdelivr.net/npm/@mediapipe/tasks-vision@0.10.2/wasm");

  

  objectDetector = await ObjectDetector.createFromOptions(vision, {

    baseOptions: {

      modelAssetPath: `https://storage.googleapis.com/mediapipe-models/object_detector/efficientdet_lite2/float16/1/efficientdet_lite2.tflite`,

      delegate: "GPU"

    },

    scoreThreshold: 0.2, 

    runningMode: runningMode

  });

  

  loader.style.display = "none";

  console.log("Model Loaded: EfficientDet-Lite2");

};



initializeObjectDetector();



// 2. NMS Filter Function

function filterDetections(detections, iouLimit) {

  detections.sort((a, b) => b.categories[0].score - a.categories[0].score);

  

  const selected = [];

  const active = new Array(detections.length).fill(true);



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

    if (!active[i]) continue;

    if (detections[i].categories[0].score < confidenceThreshold) continue;



    selected.push(detections[i]);

    const boxA = detections[i].boundingBox;



    for (let j = i + 1; j < detections.length; j++) {

      if (!active[j]) continue;

      

      const boxB = detections[j].boundingBox;

      

      const x1 = Math.max(boxA.originX, boxB.originX);

      const y1 = Math.max(boxA.originY, boxB.originY);

      const x2 = Math.min(boxA.originX + boxA.width, boxB.originX + boxB.width);

      const y2 = Math.min(boxA.originY + boxA.height, boxB.originY + boxB.height);



      if (x2 < x1 || y2 < y1) continue; 



      const intersection = (x2 - x1) * (y2 - y1);

      const areaA = boxA.width * boxA.height;

      const areaB = boxB.width * boxB.height;

      const union = areaA + areaB - intersection;

      

      const iou = intersection / union;



      if (iou > iouLimit) {

        active[j] = false; 

      }

    }

  }

  return selected;

}



// 3. Handle Clicks & Draw

async function handleClick(event) {

  if (!objectDetector) return;

  

  const img = event.target;

  const card = img.parentNode;



  // -- NEW: Inject/Show Inference Bar --

  let infoPanel = card.querySelector('.inference-panel');

  if (!infoPanel) {

    infoPanel = document.createElement('div');

    infoPanel.className = 'inference-panel';

    infoPanel.innerHTML = `

      <div class="inference-status">

        <span>Running Inference...</span>

        <span>Please wait</span>

      </div>

      <div class="progress-track">

        <div class="progress-bar"></div>

      </div>

    `;

    card.appendChild(infoPanel);

  }



  // 1. Show Loading State

  card.classList.add('processing');

  // Clear old boxes immediately so the user sees a "reset"

  card.querySelectorAll('.highlighter, .label-tag').forEach(el => el.remove());



  // 2. Force a tiny delay so the browser renders the loading bar 

  // before the heavy synchronous AI detection freezes the thread.

  await new Promise(resolve => requestAnimationFrame(() => setTimeout(resolve, 50)));

  

  try {

    // 3. Run Detection

    const predictions = objectDetector.detect(img);

    const filteredDetections = filterDetections(predictions.detections, overlapThreshold);

    displayDetections(filteredDetections, img);

  } catch(e) {

    console.error(e);

    alert("Error running model");

  } finally {

    // 4. Hide Loading State

    card.classList.remove('processing');

  }

}



function displayDetections(detections, img) {

  const ratioX = img.width / img.naturalWidth;

  const ratioY = img.height / img.naturalHeight; 

  

  detections.forEach(detection => {

    const box = detection.boundingBox;

    const category = detection.categories[0];

    const score = Math.round(category.score * 100);

    const color = getColorForLabel(category.categoryName);



    const highlighter = document.createElement("div");

    highlighter.className = "highlighter";

    highlighter.style.left = `${box.originX * ratioX}px`;

    highlighter.style.top = `${box.originY * ratioY}px`;

    highlighter.style.width = `${box.width * ratioX}px`;

    highlighter.style.height = `${box.height * ratioY}px`;

    highlighter.style.borderColor = color;

    highlighter.style.backgroundColor = color + "20"; 



    const label = document.createElement("div");

    label.className = "label-tag";

    label.innerText = `${category.categoryName} ${score}%`;

    label.style.backgroundColor = color;

    

    const topPos = (box.originY * ratioY) - 25;

    label.style.left = `${box.originX * ratioX}px`;

    label.style.top = `${topPos > 0 ? topPos : (box.originY * ratioY)}px`;



    img.parentNode.appendChild(highlighter);

    img.parentNode.appendChild(label);

  });

}



function getColorForLabel(label) {

  let hash = 0;

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

    hash = label.charCodeAt(i) + ((hash << 5) - hash);

  }

  const c = (hash & 0x00FFFFFF).toString(16).toUpperCase();

  return "#" + "00000".substring(0, 6 - c.length) + c;

}



// 4. Initialization & Event Listeners

const imageContainer = document.getElementById("imageContainer");



imageContainer.addEventListener('click', (e) => {

  if (e.target.tagName === 'IMG') handleClick(e);

});



document.getElementById('imageUpload').addEventListener('change', (e) => {

  const file = e.target.files[0];

  if (!file) return;

  

  const reader = new FileReader();

  reader.onload = (event) => {

    const div = document.createElement('div');

    div.className = 'detect-card';

    const img = document.createElement('img');

    img.src = event.target.result;

    div.appendChild(img);

    imageContainer.insertBefore(div, imageContainer.firstChild); 

  };

  reader.readAsDataURL(file);

});



document.getElementById('confidenceSlider').addEventListener('input', (e) => {

  confidenceThreshold = e.target.value / 100;

  document.getElementById('confValue').innerText = e.target.value;

});



document.getElementById('overlapSlider').addEventListener('input', (e) => {

  overlapThreshold = e.target.value / 100;

  document.getElementById('overlapValue').innerText = e.target.value;

});

</script>

</body>
</html>