Untitled

<!DOCTYPE html>
<html lang="en">
<head>
      <link rel="stylesheet" href="https://maxcdn.bootstrapcdn.com/bootstrap/3.4.0/css/bootstrap.min.css">
      <style type="text/css">
        .previous {
              background-color: #f1f1f1;
              color: black;
            }
            .round {
              border-radius: 50%;
            }
            a {
              text-decoration: none;
              display: inline-block;
              padding: 8px 16px;
            }

            a:hover {
              background-color: #ddd;
              color: black;
            }
      </style>
</head>
<body>
    <nav class="navbar navbar-inverse">
      <div class="container-fluid">
        <div class="navbar-header">
            <a href="home.html" class="previous round">&#8249;</a>
          <!-- <a class="navbar-brand" href="#">Deep Learning Course Projects</a> -->
        </div>
      </div>
    </nav>
    <div class="text-center">
        <h2>Car Brand detection</h2>
        <h5>Ahmed Adel</h5>
        <h5>Fady George</h5>
    </div>
    <h2>Problem Statment</h2>
    <h3> There are many people who are crazy about cars so
we decided to implement this idea which is detecting
car make-year-model from an image.</h3>
    <h2>Input/ output Example</h2>
    <h2>Gantt chart</h2>
    <h2>Dataset</h2>
    <h3> We used Stanford Car Dataset by classes folder which has 16185 images </h3>
    <h2>State of the art</h2>
    <h3>The best accuracy was 88% implemented by samuel
Freshman where the network trained on 49 classes
while the accuracy of GoogleNet was 80% by having
196 classes.</h3>
    <h2>Data pre-processing</h2>
    <h3>We have applied data augmentation as follows : random horizontal flips , random rotation < 30 degree , random zoom < 10 and random lighting < 10</h3>
    <h2>Models Used</h2>
    <h3>1) Yolo v3 for image segmentation</h3>
    <h3>2) ResNet34 for image detection</h3>
    <h3>3) VGG16 for image detection</h3>
    <h2>Fine Tuning apllied</h2>
    <h2>Results</h2>
    <h3>VGG16 reached validation accuracy 71%</h3>
    <h3>Fine tuning ResNet34 the model was able to reach 87% of
accuracy</h3>
</body>
</html>