English | MP4 | AVC 1280×720 | AAC 48KHz 2ch | 14 Hours | 5.04 GB
Use CNN for Image Recognition, Computer vision using TensorFlow & VGGFace2! For Data Science, Machine Learning, and AI
Convolutional Neural Networks (CNNs) are considered as game-changers in the field of computer vision, particularly after AlexNet in 2012. And the good news is CNNs are not restricted to images only. They are everywhere now, ranging from audio processing to more advanced reinforcement learning (i.e., Resnets in AlphaZero). So, the understanding of CNNs becomes almost inevitable in all the fields of Data Science. Even most of the Recurrent Neural Networks rely on CNNs these days. So, keeping all these concerns in parallel, with this course, you can take your career to the next level with an expert grip on the concepts and implementations of CNNs in Data Science.
The course ‘Mastering Convolutional Neural Networks, Theory and Practice in Python, TensorFlow 2.0’ is crafted to reflect the in-demand skills in the marketplace that will help you in mastering the concepts and methodology with regards to Python. The course is:
- Easy to understand.
- Practical with live coding.
- Rich with state-of-the-art and recently discovered CNN models by the champions in this field.
How is this course different?
This course has been designed for beginners. However, we will go far deep gradually.
Also, this course is a quick compilation of all the basics, and it encourages you to press forward and experience more than what you have learned. By the end of every module, you will work on the assigned Homework/tasks/activities, which will evaluate / (further build) your learning based on the previous concepts and methods. Several of these activities will be coding-based to get you up and running with implementations.
Data Science is certainly a rewarding career that not only allows you to solve some of the most interesting problems, but also offers you a handsome salary package. With a core understanding of CNNs, you can back up your business and ensure emerging career growth.
What you’ll learn
- The importance of Convolutional Neural Networks (CNNs) in Data Science.
- The reasons to shift from hand engineering (classical computer vision) to CNNs.
- The essential concepts from the absolute beginning with comprehensive unfolding with examples in Python.
- Practical explanation and live coding with Python.
- An overview of concepts of Deep Learning theory.
- Evolutions of CNNs from LeNet (1990s) to MobileNets (2020s).
- Deep details of CNNs with examples of training CNNs from scratch.
- TensorFlow (Deep learning framework by Google).
- The use and applications of state-of-the-art CNNs (with implementations in state-of-the-art framework TensorFlow) that are much
- more recent and advanced in terms of accuracy and efficiency.
- The use and applications of state-of-the-art pre-trained CNNs (with implementations in state-of-the-art framework TensorFlow) for transfer learning on your own dataset.
- Building your own applications for human Face Verification and Neural Style Transfer.
1 Instructor Introduction
2 Why CNN
3 Focus of the Course
4 Request for Your Honest Review
5 Link to Github to get the Python Notebooks
6 Gray Scale Images
7 RGB Images
8 Reading and Showing Images in Python
9 Converting an Image to Grayscale in Python
10 Image Formation
11 Image Blurring 1
12 Image Blurring 2
13 General Image Filtering
15 Edge Detection
16 Image Sharpening
17 Implementation of Image Blurring Edge Detection Image Sharpening in Python
18 Parameteric Shape Detection
19 Image Processing Activity
20 Introduction to Object Detection
21 Classification PipleLine
22 Sliding Window Implementation
23 Shift Scale Rotation Invariance
24 Person Detection
25 HOG Features
26 Hand Engineering vs CNNs
27 Object Detection Activity
Deep Neural Network Overview
28 Neuron and Perceptron
29 DNN Architecture
30 FeedForward FullyConnected MLP
31 Calculating Number of Weights of DNN
32 Number of Nuerons vs Number of Layers
33 Discriminative vs Generative Learning
34 Universal Approximation Therorem
35 Why Depth
36 Decision Boundary in DNN
38 Activation Function
39 DNN Training Parameters
40 Gradient Descent
42 Training DNN Animantion
43 Weigth Initialization
44 Batch miniBatch Stocastic Gradient Descent
45 Batch Normalization
46 Rprop and Momentum
47 Convergence Animation
48 DropOut, Early Stopping and Hyperparameters
Deep Neural Network Architecture
49 Convolution Revisited
50 Implementing Convolution in Python Revisited
51 Why Convolution
52 Filters Padding Strides
53 Pooling Tensors
54 CNN Example
55 Convolution and Pooling Details
56 NonVectorized Implementations of Conv2d and Pool2d
57 Deep Neural Network Architecture Activity
Gradient Descent in CNNs
58 Example Setup
59 Why Derivaties
60 What is Chain Rule
61 Applying Chain Rule
62 Gradients of MaxPooling Layer
63 Gradients of Convolutional Layer
64 Extending To Multiple Filters
65 Extending to Multiple Layers
66 Implementation in Numpy ForwardPass
67 Implementation in Numpy BackwardPass 1
68 Implementation in Numpy BackwardPass 2
69 Implementation in Numpy BackwardPass 3
70 Implementation in Numpy BackwardPass 4
71 Implementation in Numpy BackwardPass 5
72 Gradient Descent in CNNs Activity
Introduction to TensorFlow
74 FashionMNIST Example Plan Neural Network
75 FashionMNIST Example CNN
76 Introduction to TensorFlow Activity
83 Classical CNNs Activity
84 What is Transfer learning
85 Why Transfer Learning
86 ImageNet Challenge
87 Practical Tips
88 Project in TensorFlow
89 Transfer Learning Activity
90 Image Classfication Revisited
91 Sliding Window Object Localization
92 Sliding Window Efficient Implementation
93 Yolo Introduction
94 Yolo Training Data Generation
95 Yolo Anchor Boxes
96 Yolo Algorithm
97 Yolo Non Maxima Supression
99 Yolo Activity
100 Problem Setup
101 Project Implementation
102 Face Verification Activity
Neural Style Transfer
103 Problem Setup
104 Implementation Tensorflow Hub
105 THANK YOU Bonus Video