Gain insights into image data processing and CNNs for image classification using TensorFlow. Delve into CNN architectures and their applications, requiring Python and TensorFlow knowledge.

irml.tar.gz

image-rec

If you want to dive into the technology behind computer vision and self driving cars, this is where to start.

In this course you'll learn how to process data from image files and create convolutional neural networks (CNNs) to classify different types of images. After completing this course, you will have a solid understanding of why CNNs work and which CNN architectures to use for different tasks.

The code for this course is built around the TensorFlow framework, one of the premier frameworks for industry machine learning, and the Python pandas library for data analysis. Knowledge of Python and TensorFlow are prerequisites. 

This course was created by AdaptiLab, a company specializing in evaluating, sourcing, and upskilling enterprise machine learning talent. It is built in collaboration with industry machine learning experts from Google, Microsoft, Amazon, and Apple.

Image Recognition with Machine Learning

<p>In this section of the course you will be writing a basic convolutional neural network (CNN) for handwritten digit recognition of the MNIST dataset. We will take an in-depth look at a model based off of Yann LeCun's <a href="http://yann.lecun.com/exdb/lenet/" target="_blank">LeNet-5</a>.</p>



# A. Digit recognition

When we start building an image recognition model for the first time, it's usually a good idea to train and evaluate it on a relatively simple dataset. This allows for less complicated models, quicker training, and less image preprocessing. Basically, we want to learn the inner workings of a convolutional neural network before we even try to apply it to more complex tasks.

One of the simplest tasks we can perform is handwritten digit recognition. Given an image of a handwritten digit (i.e., 0, 1, ..., 9), we want our model to be able to correctly classify its numeric value. Though this task seems relatively simple, it is actually used fairly often in real life, such as automatically extracting credit card numbers from a picture. The dataset we will use for digit recognition is the <a href="https://en.wikipedia.org/wiki/MNIST_database" target="_blank">MNIST dataset</a>, which is the de facto dataset used for machine learning-based digit recognition.


# B. CNN dominance

<div id="section2_l0" class="collapse show">
<p>Nowadays, every task in the field of image recognition is dominated by convolutional neural networks (CNNs). Since the introduction of LeNet-5, the pioneering CNN, in 1998 for digit recognition, CNN architectures have become king in nearly all computer vision tasks. The CNN forms the backbone of many of today's cutting-edge technologies, such as self-driving cars, robotics, and facial recognition.</p>
<p>Although CNNs dominate image related tasks, we could still theoretically use a multilayer perceptron model for classifying images. However, the MLP would not perform nearly as well as a CNN. So what exactly makes a CNN special for image input data? The key to the model's success is its <i>convolutional layers</i>, which is one of the main topics we'll discuss in the following chapters.</p>
</div>


# B. CNN dominance

<div id="section2_l0">
<p>Nowadays, every task in the field of image recognition is dominated by convolutional neural networks (CNNs). Since the introduction of LeNet-5, the pioneering CNN, in 1998 for digit recognition, CNN architectures have become king in nearly all computer vision tasks. The CNN forms the backbone of many of today's cutting-edge technologies, such as self-driving cars, robotics, and facial recognition.</p>
<p>Although CNNs dominate image related tasks, we could still theoretically use a multilayer perceptron model for classifying images. However, the MLP would not perform nearly as well as a CNN. So what exactly makes a CNN special for image input data? The key to the model's success is its <i>convolutional layers</i>, which is one of the main topics we'll discuss in the following chapters.</p>
</div>


What you'll learn from this course

Image Processing

CNN

SqueezeNet

ResNet

Introduction

A. Digit recognition

B. CNN dominance