-
Notifications
You must be signed in to change notification settings - Fork 0
/
project_details_CTS.html
78 lines (61 loc) · 5.01 KB
/
project_details_CTS.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
<!DOCTYPE HTML>
<html>
<head>
<title>Classifying Traffic Sign - Project Details</title>
<meta charset="utf-8" />
<meta name="viewport" content="width=device-width, initial-scale=1, user-scalable=no" />
<link rel="stylesheet" href="assets/css/main.css" />
</head>
<body class="is-preload">
<article class="wrapper style1">
<div class="container">
<div class="row">
<div class="col-12">
<header>
<h2>Future Mobility Challenge: Deep Learning and Computer Vision for Autonomous Vehicles (AVs)</h2>
<caption>Challenge Mentor: Ryan Ahmed, Ph.D., MBA Professor, McMaster University</caption>
</header>
<h2><a href="https://github.com/arol9204/CLASSIFY_TRAFFIC_SIGNS" target= "_blank" class="icon brands fa-github"><span class="label">Github</span></a></h2>
<section>
<h3>Project Details</h3>
<p>The automotive industry is going through a paradigm shift from conventional, human-driven vehicles into Autonomous,Artificial Intelligence (AI)-powered vehicles. AVs offer a safe, efficient, and cost-effective solution that will dramaticallyredefine the future of human mobility. AVs are expected to save over half a million lives and generate enormous economicopportunities in excess of $1 trillion dollars by 2035. From GM’s Cruise and Google’s autonomous car (Waymo) to Uber’sself-driving car-sharing service, the automotive industry is on a billion-dollar quest to deploy the most technologicallyadvanced vehicles on the road.</p>
<p>In this challenge, you are tasked to train and evaluate several deep CNNs to classify traffic sign images.
The dataset contains colored 32 x 32 pixels images with 43 different traffic sign classes.
Images are divided into 34799 images for training, 12630 for testing and 4410 for validation.
</p>
<div class="image fit">
<img src="images/traffic signs.png" alt="Classifying Traffic Sign" />
</div>
<ol>
<h4> Challenge Deliverables: </h4>
<li>Research AlexNet Convolutional Neural Network, draw the network architecture, and indicate the use of each of the layers such as convolution/feature extraction, max pooling, and flattening layers. </li>
<li>Visualize a sample image from each class in the dataset. Display the corresponding label and perform a sanity check. </li>
<li>Train a CNN based on AlexNet to classify traffic signs. Train the model using 15 epochs and plot the network accuracy vs. the number of epochs. </li>
<li>Test the trained CNN and evaluate its performance on the testing (holdout) dataset. Note that the testing dataset has never been seen by the model during training.</li>
<li>Fine-tune the trained model by training it on additional datasets. Feel free to collect data from the internet or using a mobile device.</li>
<li>Explore 3 different ways to improve the network accuracy such as increasing the number of epochs, adding additional convolutional/max-pooling layers, or adding dropout layers. Retrain the network and display the results. Any reasonable answer should be sufficient.</li>
<li>Build 3 different deep convolutional neural networks with various architectures. Feel free to choose the number of layers, number of kernels (feature detectors), number of dense (Fully-connected) layers, number of neurons in the dense layer, activation functions, regularization such as dropout..etc. Retrain the network and plot the results.</li>
<li>Test the trained deep CNNs on brand-new images. This might involve collecting traffic sign data and evaluating the system's performance and make adjustment as necessary.</li>
</ol>
</section>
<section>
<h3>Additional Content</h3>
<p>1st Place Prize (Value): $500</p>
<div class="image fit">
<img src="images/Future Mobility Challenge 2023.jpeg" alt="Classifying Traffic Sign" />
<caption>*I'm the third person from the right to the left.</caption>
</div>
</section>
<p><a href="index.html">< Back to Portfolio</a></p>
</div>
</div>
</div>
</article>
<script src="assets/js/jquery.min.js"></script>
<script src="assets/js/jquery.scrolly.min.js"></script>
<script src="assets/js/browser.min.js"></script>
<script src="assets/js/breakpoints.min.js"></script>
<script src="assets/js/util.js"></script>
<script src="assets/js/main.js"></script>
</body>
</html>