title
Convolutional Neural Network (CNN) | Convolutional Neural Networks With TensorFlow | Edureka

description
The code referenced in this video is from https://YouTube.com/Sentdex and https://pythonprogramming.net/convolutional-neural-network-kats-vs-dogs-machine-learning-tutorial/ πŸ”₯ TensorFlow Training (Use Code "π˜πŽπ”π“π”ππ„πŸπŸŽ") - https://www.edureka.co/ai-deep-learning-with-tensorflow This Edureka "Convolutional Neural Network Tutorial" video (Blog: https://goo.gl/4zxMfU) will help you in understanding what is Convolutional Neural Network and how it works. It also includes a use-case, in which we will be creating a classifier using TensorFlow. Below are the topics covered in this tutorial: 1. How a Computer Reads an Image? 2. Why can't we use Fully Connected Networks for Image Recognition? 3. What is Convolutional Neural Network? 4. How Convolutional Neural Networks Work? 5. Use-Case (dog and cat classifier) Subscribe to our channel to get video updates. Hit the subscribe button above. Check our complete Deep Learning With TensorFlow playlist here: https://goo.gl/cck4hE PG in Artificial Intelligence and Machine Learning with NIT Warangal : https://www.edureka.co/post-graduate/machine-learning-and-ai Post Graduate Certification in Data Science with IIT Guwahati - https://www.edureka.co/post-graduate/data-science-program (450+ Hrs || 9 Months || 20+ Projects & 100+ Case studies) - - - - - - - - - - - - - - How it Works? 1. This is 21 hrs of Online Live Instructor-led course. Weekend class: 7 sessions of 3 hours each. 2. We have a 24x7 One-on-One LIVE Technical Support to help you with any problems you might face or any clarifications you may require during the course. 3. At the end of the training you will have to undergo a 2-hour LIVE Practical Exam based on which we will provide you a Grade and a Verifiable Certificate! - - - - - - - - - - - - - - About the Course Edureka's Deep learning with Tensorflow course will help you to learn the basic concepts of TensorFlow, the main functions, operations and the execution pipeline. Starting with a simple β€œHello Word” example, throughout the course you will be able to see how TensorFlow can be used in curve fitting, regression, classification and minimization of error functions. This concept is then explored in the Deep Learning world. You will evaluate the common, and not so common, deep neural networks and see how these can be exploited in the real world with complex raw data using TensorFlow. In addition, you will learn how to apply TensorFlow for backpropagation to tune the weights and biases while the Neural Networks are being trained. Finally, the course covers different types of Deep Architectures, such as Convolutional Networks, Recurrent Networks and Autoencoders. Delve into neural networks, implement Deep Learning algorithms, and explore layers of data abstraction with the help of this Deep Learning with TensorFlow course. - - - - - - - - - - - - - - Who should go for this course? The following professionals can go for this course: 1. Developers aspiring to be a 'Data Scientist' 2. Analytics Managers who are leading a team of analysts 3. Business Analysts who want to understand Deep Learning (ML) Techniques 4. Information Architects who want to gain expertise in Predictive Analytics 5. Professionals who want to captivate and analyze Big Data 6. Analysts wanting to understand Data Science methodologies However, Deep learning is not just focused to one particular industry or skill set, it can be used by anyone to enhance their portfolio. - - - - - - - - - - - - - - Why Learn Deep Learning With TensorFlow? TensorFlow is one of the best libraries to implement Deep Learning. TensorFlow is a software library for numerical computation of mathematical expressions, using data flow graphs. Nodes in the graph represent mathematical operations, while the edges represent the multidimensional data arrays (tensors) that flow between them. It was created by Google and tailored for Machine Learning. In fact, it is being widely used to develop solutions with Deep Learning. Machine learning is one of the fastest-growing and most exciting fields out there, and Deep Learning represents its true bleeding edge. Deep learning is primarily a study of multi-layered neural networks, spanning over a vast range of model architectures. Traditional neural networks relied on shallow nets, composed of one input, one hidden layer and one output layer. Deep-learning networks are distinguished from these ordinary neural networks having more hidden layers, or so-called more depth. These kinds of nets are capable of discovering hidden structures within unlabeled and unstructured data (i.e. images, sound, and text), which constitutes the vast majority of data in the world. For more information, please write back to us at sales@edureka.co or call us at IND: 9606058406 / US: 18338555775 (toll-free). Facebook: https://www.facebook.com/edurekaIN/ Twitter: https://twitter.com/edurekain LinkedIn: https://www.linkedin.com/company/edureka

detail
{'title': 'Convolutional Neural Network (CNN) | Convolutional Neural Networks With TensorFlow | Edureka', 'heatmap': [{'end': 86.879, 'start': 50.422, 'weight': 0.744}, {'end': 136.896, 'start': 102.86, 'weight': 0.819}, {'end': 361.964, 'start': 318.021, 'weight': 0.712}, {'end': 410.6, 'start': 373.253, 'weight': 0.764}, {'end': 454.137, 'start': 424.128, 'weight': 0.742}, {'end': 640.814, 'start': 599.916, 'weight': 0.724}, {'end': 776.837, 'start': 744.526, 'weight': 0.909}, {'end': 1097.25, 'start': 1080.484, 'weight': 0.701}], 'summary': 'Covers the limitations of fully connected networks for image recognition, the need for convolutional neural networks, and practical applications such as classifying images of x and o, achieving a 91% match for classifying x over o, and training a model to classify dog and cat images with 88% accuracy and a loss of 0.2973 after 10 epochs.', 'chapters': [{'end': 177.454, 'segs': [{'end': 58.204, 'src': 'embed', 'start': 29.133, 'weight': 0, 'content': [{'end': 34.576, 'text': "So when a human will see this image, he'll first notice there are a lot of buildings in different colors and stuff like that.", 'start': 29.133, 'duration': 5.443}, {'end': 36.797, 'text': 'But how a computer will read this image?', 'start': 35.036, 'duration': 1.761}, {'end': 44.86, 'text': 'So, basically, there will be three channels one will be red, another will be green and finally we have blue channel, which is popularly known as RGB.', 'start': 37.338, 'duration': 7.522}, {'end': 49.901, 'text': 'So each of these channels will have their own respective pixel values, as you can see it over here.', 'start': 45.3, 'duration': 4.601}, {'end': 58.204, 'text': 'So when I say that image size is B cross A cross three, it means that there are B rows, A columns, and three channels.', 'start': 50.422, 'duration': 7.782}], 'summary': 'Explanation of how a computer reads an image using rgb channels and pixel values.', 'duration': 29.071, 'max_score': 29.133, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A29133.jpg'}, {'end': 86.879, 'src': 'heatmap', 'start': 50.422, 'weight': 0.744, 'content': [{'end': 58.204, 'text': 'So when I say that image size is B cross A cross three, it means that there are B rows, A columns, and three channels.', 'start': 50.422, 'duration': 7.782}, {'end': 66.026, 'text': 'So if somebody tells you that the size of an image is 28 cross 28 cross three pixels, it means that it has 28 rows, 28 columns, and three channels.', 'start': 58.364, 'duration': 7.662}, {'end': 70.629, 'text': 'So this is how a computer sees an image and this is for colored images for black and white images.', 'start': 66.306, 'duration': 4.323}, {'end': 71.61, 'text': 'We have only two channels.', 'start': 70.649, 'duration': 0.961}, {'end': 76.092, 'text': "So let's move forward and we'll see why can't we use fully connected networks for image classification.", 'start': 72.05, 'duration': 4.042}, {'end': 79.975, 'text': 'So consider an image which has 28 cross 28 cross 3 pixels.', 'start': 76.533, 'duration': 3.442}, {'end': 86.879, 'text': 'So when I feed in this image to a fully connected network like this then the total number of weights required in the first hidden layer will be 2352.', 'start': 80.335, 'duration': 6.544}], 'summary': 'Image size is represented as bxax3, 28x28x3 pixels requires 2352 weights in first layer.', 'duration': 36.457, 'max_score': 50.422, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A50422.jpg'}, {'end': 142.4, 'src': 'heatmap', 'start': 97.998, 'weight': 1, 'content': [{'end': 102.86, 'text': 'So if I take an image which has 200 cross, 200 cross three pixels and I feed it to a fully connected network,', 'start': 97.998, 'duration': 4.862}, {'end': 107.362, 'text': 'so at that time the number of weights required at the first hidden layer itself will be 120, 000 guys.', 'start': 102.86, 'duration': 4.502}, {'end': 112.263, 'text': 'So we need to deal with such huge amount of parameters and obviously we require more number of neurons.', 'start': 107.642, 'duration': 4.621}, {'end': 114.664, 'text': 'So that can eventually lead to overfitting.', 'start': 112.544, 'duration': 2.12}, {'end': 118.846, 'text': "So that's why we cannot use fully connected network for image classification.", 'start': 114.884, 'duration': 3.962}, {'end': 121.347, 'text': "Now let's see why we need convolutional neural networks.", 'start': 119.286, 'duration': 2.061}, {'end': 128.431, 'text': 'So basically in convolutional neural network a neuron in the layer will only be connected to a small region of the layer before it.', 'start': 121.847, 'duration': 6.584}, {'end': 132.454, 'text': "So if you consider this particular neuron which I'm highlighting right now is only connected to three other neurons.", 'start': 128.511, 'duration': 3.943}, {'end': 136.896, 'text': 'Unlike the fully connected network where this particular neuron will be connected to all these five neurons.', 'start': 132.914, 'duration': 3.982}, {'end': 142.4, 'text': 'Because of this we need to handle less amount of weights and in turn we need less number of neurons as well.', 'start': 137.276, 'duration': 5.124}], 'summary': 'Fully connected networks require 120,000 weights for 200x200x3 image, leading to overfitting. convolutional neural networks use fewer weights and neurons for image classification.', 'duration': 44.402, 'max_score': 97.998, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A97998.jpg'}, {'end': 173.191, 'src': 'embed', 'start': 145.742, 'weight': 2, 'content': [{'end': 153.005, 'text': 'So convolutional neural networks are special type of feed forward artificial neural networks which is inspired from visual cortex.', 'start': 145.742, 'duration': 7.263}, {'end': 158.726, 'text': 'So visual cortex is nothing but a small region in our brain which is present somewhere here where you can see the bulb.', 'start': 153.065, 'duration': 5.661}, {'end': 160.507, 'text': 'and basically what happened?', 'start': 158.726, 'duration': 1.781}, {'end': 167.549, 'text': 'there was an experiment conducted and people got to know that visual cortex is small regions of cells that are sensitive to specific regions of visual field.', 'start': 160.507, 'duration': 7.042}, {'end': 173.191, 'text': 'So what I mean by that is for example some neurons in the visual cortex fires when exposed to vertical edges.', 'start': 167.769, 'duration': 5.422}], 'summary': 'Convolutional neural networks are inspired by the visual cortex, where specific neurons are sensitive to particular visual features.', 'duration': 27.449, 'max_score': 145.742, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A145742.jpg'}], 'start': 0.009, 'title': 'Convolutional neural networks', 'summary': 'Explains the limitations of fully connected networks for image recognition, the need for convolutional neural networks, which require less weights and neurons, inspired by the visual cortex.', 'chapters': [{'end': 177.454, 'start': 0.009, 'title': 'Convolutional neural networks', 'summary': 'Explains how a computer reads an image using channels, the limitations of fully connected networks for image recognition due to the large number of weights, and the need for convolutional neural networks which are inspired by the visual cortex and require less weights and neurons.', 'duration': 177.445, 'highlights': ['A computer reads an image using three channels - red, green, and blue (RGB) - with each channel having its own respective pixel values, for colored images, and two channels for black and white images. Describes how a computer reads an image using channels and pixel values, providing insight into the RGB system and the differences for colored and black and white images.', 'Fully connected networks are not suitable for image classification due to the large number of weights required, which increases significantly for larger images, leading to overfitting. Explains the limitations of fully connected networks for image classification, citing the exponential increase in weights for larger images and the resulting overfitting issues.', 'Convolutional neural networks are inspired by the visual cortex and have neurons connected to a small region of the layer before it, requiring fewer weights and less number of neurons compared to fully connected networks. Highlights the inspiration behind convolutional neural networks from the visual cortex and the advantage of requiring fewer weights and neurons due to the connection to a small region of the layer before it.']}], 'duration': 177.445, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A9.jpg', 'highlights': ['Describes how a computer reads an image using channels and pixel values, providing insight into the RGB system and the differences for colored and black and white images.', 'Explains the limitations of fully connected networks for image classification, citing the exponential increase in weights for larger images and the resulting overfitting issues.', 'Highlights the inspiration behind convolutional neural networks from the visual cortex and the advantage of requiring fewer weights and neurons due to the connection to a small region of the layer before it.']}, {'end': 508.856, 'segs': [{'end': 210.99, 'src': 'embed', 'start': 177.734, 'weight': 0, 'content': [{'end': 180.797, 'text': 'And that is nothing but the motivation behind convolutional neural network.', 'start': 177.734, 'duration': 3.063}, {'end': 184.06, 'text': 'So now let us understand how exactly a convolutional neural network works.', 'start': 181.037, 'duration': 3.023}, {'end': 186.683, 'text': 'So generally a convolutional neural network has three layers.', 'start': 184.361, 'duration': 2.322}, {'end': 189.826, 'text': 'Convolutional layer, ReLU layer, pooling layer and fully connected layer.', 'start': 186.783, 'duration': 3.043}, {'end': 191.848, 'text': "We'll understand each of these layers one by one.", 'start': 189.846, 'duration': 2.002}, {'end': 197.158, 'text': "We'll take an example of a classifier but that can classify an image of an X as well as an O.", 'start': 192.168, 'duration': 4.99}, {'end': 199.88, 'text': "So with this example, we'll be understanding all these four layers.", 'start': 197.158, 'duration': 2.722}, {'end': 201.081, 'text': "So let's begin guys.", 'start': 200.22, 'duration': 0.861}, {'end': 202.963, 'text': 'Now there are certain trickier cases.', 'start': 201.541, 'duration': 1.422}, {'end': 209.749, 'text': 'So what I mean by that is X can be represented in these four forms as well, right? So these are nothing but the deformed images of X.', 'start': 202.983, 'duration': 6.766}, {'end': 210.99, 'text': 'Similarly for O as well.', 'start': 209.749, 'duration': 1.241}], 'summary': 'Introduction to convolutional neural network with focus on its layers and an example of classifying x and o images.', 'duration': 33.256, 'max_score': 177.734, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A177734.jpg'}, {'end': 361.964, 'src': 'heatmap', 'start': 318.021, 'weight': 0.712, 'content': [{'end': 320.282, 'text': "So we'll be considering these three features or filters.", 'start': 318.021, 'duration': 2.261}, {'end': 324.003, 'text': 'This is a diagonal filter, this is again a diagonal filter, and this is nothing but a small image.', 'start': 320.482, 'duration': 3.521}, {'end': 326.804, 'text': "So we'll take these three filters and we'll move forward.", 'start': 324.363, 'duration': 2.441}, {'end': 332.246, 'text': 'So what we are going to do is we are going to compare these features, the small pieces of the bigger image.', 'start': 327.164, 'duration': 5.082}, {'end': 337.008, 'text': 'We are going to put it on the input image and if it matches, then the image will be classified correctly.', 'start': 332.606, 'duration': 4.402}, {'end': 338.088, 'text': "Now we'll begin guys.", 'start': 337.328, 'duration': 0.76}, {'end': 339.689, 'text': 'The first layer is convolution layer.', 'start': 338.128, 'duration': 1.561}, {'end': 342.13, 'text': 'So these are the beginning two steps of this particular layer.', 'start': 339.869, 'duration': 2.261}, {'end': 346.331, 'text': 'First we need to line up the feature and the image and then multiply image by the corresponding feature pixel.', 'start': 342.39, 'duration': 3.941}, {'end': 347.991, 'text': 'Now let me explain you with an example.', 'start': 346.63, 'duration': 1.361}, {'end': 350.494, 'text': "So this is our first diagonal feature that we'll take.", 'start': 348.292, 'duration': 2.202}, {'end': 358, 'text': 'We are going to put this particular feature on our image of X, all right? And we are going to multiply the corresponding pixel value.', 'start': 350.754, 'duration': 7.246}, {'end': 361.964, 'text': "So one will be multiplied with one, we'll get one and we'll put it in another matrix.", 'start': 358.04, 'duration': 3.924}], 'summary': 'The convolution layer compares features with input image to classify correctly.', 'duration': 43.943, 'max_score': 318.021, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A318021.jpg'}, {'end': 358, 'src': 'embed', 'start': 324.363, 'weight': 2, 'content': [{'end': 326.804, 'text': "So we'll take these three filters and we'll move forward.", 'start': 324.363, 'duration': 2.441}, {'end': 332.246, 'text': 'So what we are going to do is we are going to compare these features, the small pieces of the bigger image.', 'start': 327.164, 'duration': 5.082}, {'end': 337.008, 'text': 'We are going to put it on the input image and if it matches, then the image will be classified correctly.', 'start': 332.606, 'duration': 4.402}, {'end': 338.088, 'text': "Now we'll begin guys.", 'start': 337.328, 'duration': 0.76}, {'end': 339.689, 'text': 'The first layer is convolution layer.', 'start': 338.128, 'duration': 1.561}, {'end': 342.13, 'text': 'So these are the beginning two steps of this particular layer.', 'start': 339.869, 'duration': 2.261}, {'end': 346.331, 'text': 'First we need to line up the feature and the image and then multiply image by the corresponding feature pixel.', 'start': 342.39, 'duration': 3.941}, {'end': 347.991, 'text': 'Now let me explain you with an example.', 'start': 346.63, 'duration': 1.361}, {'end': 350.494, 'text': "So this is our first diagonal feature that we'll take.", 'start': 348.292, 'duration': 2.202}, {'end': 358, 'text': 'We are going to put this particular feature on our image of X, all right? And we are going to multiply the corresponding pixel value.', 'start': 350.754, 'duration': 7.246}], 'summary': 'Comparing features and classifying images using convolution layer.', 'duration': 33.637, 'max_score': 324.363, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A324363.jpg'}, {'end': 410.6, 'src': 'heatmap', 'start': 373.253, 'weight': 0.764, 'content': [{'end': 376.336, 'text': 'So we are going to complete this whole process when we are going to finish up this Matrix.', 'start': 373.253, 'duration': 3.083}, {'end': 382.781, 'text': 'All right, and once we are done finishing up the multiplication of all the corresponding pixels in the feature as well as in the image.', 'start': 376.596, 'duration': 6.185}, {'end': 384.282, 'text': 'We need to follow two more steps.', 'start': 383.061, 'duration': 1.221}, {'end': 387.685, 'text': 'We need to add them up and divide by the total number of the pixels in the feature.', 'start': 384.522, 'duration': 3.163}, {'end': 393.769, 'text': 'So what I mean by that is, after the multiplication of the corresponding pixel values, what we do?', 'start': 387.985, 'duration': 5.784}, {'end': 398.352, 'text': 'we add all these values, we divide it by the total number of pixels and we get some value right?', 'start': 393.769, 'duration': 4.583}, {'end': 403.275, 'text': 'And then now our next step is to create a map and put the value of the filter at that particular place.', 'start': 398.612, 'duration': 4.663}, {'end': 410.6, 'text': 'We saw that after multiplying the pixel value of a feature with the corresponding pixel value of that of our image, we get the output which is one.', 'start': 403.756, 'duration': 6.844}], 'summary': 'Process completion: multiply, add, divide, map, output=1', 'duration': 37.347, 'max_score': 373.253, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A373253.jpg'}, {'end': 454.137, 'src': 'heatmap', 'start': 424.128, 'weight': 0.742, 'content': [{'end': 431.011, 'text': "So yeah, this is one more example where I've moved my filter in between and after doing that, I've got the output something like this 1, 1,", 'start': 424.128, 'duration': 6.883}, {'end': 432.952, 'text': 'minus 1 and all so over here.', 'start': 431.011, 'duration': 1.941}, {'end': 438.447, 'text': "If you notice I've got couple of times minus 1 as well due to which my output that comes is 0.55.", 'start': 432.992, 'duration': 5.455}, {'end': 440.628, 'text': "Right, so I'm going to place 0.55 here.", 'start': 438.447, 'duration': 2.181}, {'end': 444.691, 'text': 'Similarly after moving the pixel after moving the filter throughout the image.', 'start': 441.029, 'duration': 3.662}, {'end': 446.272, 'text': 'I got this particular Matrix.', 'start': 444.791, 'duration': 1.481}, {'end': 452.916, 'text': 'All right, and this is for one particular feature after performing the same process for the other two filters as well.', 'start': 446.292, 'duration': 6.624}, {'end': 454.137, 'text': "I've got these two values.", 'start': 453.016, 'duration': 1.121}], 'summary': 'Filter adjustment resulted in 0.55 output, and 2 values from other filters.', 'duration': 30.009, 'max_score': 424.128, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A424128.jpg'}, {'end': 487.301, 'src': 'embed', 'start': 457.819, 'weight': 3, 'content': [{'end': 460.401, 'text': 'Let me give you a quick recap of what happens in convolution layer.', 'start': 457.819, 'duration': 2.582}, {'end': 466.105, 'text': 'So, basically, we have taken three features, all right, and one by one will take one feature, move it through the entire image,', 'start': 460.701, 'duration': 5.404}, {'end': 472.17, 'text': 'and when we are moving it at that time, we are multiplying the pixel value of the image with that of the corresponding pixel value of the filter,', 'start': 466.105, 'duration': 6.065}, {'end': 475.753, 'text': 'adding them up, dividing by the total number of pixels to get the output.', 'start': 472.17, 'duration': 3.583}, {'end': 479.656, 'text': 'So when we do that for all the filters we get we got these three outputs.', 'start': 476.133, 'duration': 3.523}, {'end': 479.936, 'text': 'All right.', 'start': 479.696, 'duration': 0.24}, {'end': 482.718, 'text': "So let's move forward and we'll see what happens in ReLU layer.", 'start': 480.256, 'duration': 2.462}, {'end': 487.301, 'text': 'So this is ReLU layer guys and people who have gone through the previous tutorial actually know what it is.', 'start': 483.058, 'duration': 4.243}], 'summary': 'Convolution layer processes three features by moving filters through the image, resulting in three outputs. relu layer is then introduced.', 'duration': 29.482, 'max_score': 457.819, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A457819.jpg'}, {'end': 519.384, 'src': 'embed', 'start': 489.783, 'weight': 4, 'content': [{'end': 491.824, 'text': 'So ReLU is nothing but a activation function.', 'start': 489.783, 'duration': 2.041}, {'end': 492.304, 'text': 'All right.', 'start': 492.064, 'duration': 0.24}, {'end': 500.65, 'text': 'So what I mean by that is it will only activate a node if the input is above a certain quantity while the input is below zero the output is also zero.', 'start': 492.324, 'duration': 8.326}, {'end': 507.335, 'text': 'All right, and when the input rises above the certain threshold, it has a linear relationship with the dependent variable.', 'start': 500.95, 'duration': 6.385}, {'end': 508.856, 'text': "Now, I'll explain you with an example.", 'start': 507.615, 'duration': 1.241}, {'end': 510.877, 'text': 'We have a graph of Rayleigh function here.', 'start': 509.396, 'duration': 1.481}, {'end': 519.384, 'text': 'So my function says that when f of x is equal to 0, if x is less than 0, and it is equal to x when x is greater than 0, all right?', 'start': 511.258, 'duration': 8.126}], 'summary': 'Relu is an activation function that outputs zero for input below zero, and linearly increases for input above zero.', 'duration': 29.601, 'max_score': 489.783, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A489783.jpg'}], 'start': 177.734, 'title': 'Convolutional neural networks', 'summary': 'Introduces the concept of convolutional neural networks, explaining its layers and how they are used to classify deformed images of x and o based on pixel values and feature matching, and explains the process of applying convolution and relu layers, resulting in three output values after passing through the convolution layer, and introduces the relu layer as an activation function.', 'chapters': [{'end': 342.13, 'start': 177.734, 'title': 'Convolutional neural networks', 'summary': 'Introduces the concept of convolutional neural networks, explaining its three layers: convolutional layer, relu layer, pooling layer, and fully connected layer, and how they are used to classify deformed images of x and o based on pixel values and feature matching.', 'duration': 164.396, 'highlights': ['Convolutional neural network has three layers: Convolutional layer, ReLU layer, pooling layer, and fully connected layer The chapter introduces the concept of convolutional neural networks and explains its three layers: convolutional layer, ReLU layer, pooling layer, and fully connected layer.', 'Classifying deformed images of X and O based on pixel values and feature matching The chapter discusses the challenge of classifying deformed images of X and O based on pixel values, and explains how convolutional neural networks use feature matching to address this challenge.', 'Explaining the process of comparing features and using filters to classify images correctly The chapter explains the process of comparing features and using filters to classify images correctly, demonstrating the use of filters to match features in deformed images for accurate classification.']}, {'end': 508.856, 'start': 342.39, 'title': 'Convolutional neural network', 'summary': 'Explains the process of applying convolution and relu layers in a convolutional neural network, resulting in three output values after passing through the convolution layer and introduces the relu layer as an activation function.', 'duration': 166.466, 'highlights': ['The process of applying convolution involves multiplying the pixel values of the image with the corresponding pixel values of the filter, adding them up, and dividing by the total number of pixels to obtain the output, resulting in three output values after passing through the convolution layer. The convolution process involves multiplying pixel values of the image with corresponding pixel values of the filter, adding them up, and dividing by the total number of pixels to obtain the output, resulting in three output values after passing through the convolution layer.', 'The ReLU layer is introduced as an activation function, where it activates a node if the input is above a certain quantity, and outputs zero when the input is below zero, while having a linear relationship with the dependent variable when the input rises above the certain threshold. ReLU layer serves as an activation function, activating a node if the input is above a certain quantity, outputting zero when the input is below zero, and having a linear relationship with the dependent variable when the input rises above the certain threshold.']}], 'duration': 331.122, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A177734.jpg', 'highlights': ['The chapter introduces the concept of convolutional neural networks and explains its three layers: convolutional layer, ReLU layer, pooling layer, and fully connected layer.', 'The chapter discusses the challenge of classifying deformed images of X and O based on pixel values, and explains how convolutional neural networks use feature matching to address this challenge.', 'The chapter explains the process of comparing features and using filters to classify images correctly, demonstrating the use of filters to match features in deformed images for accurate classification.', 'The convolution process involves multiplying pixel values of the image with corresponding pixel values of the filter, adding them up, and dividing by the total number of pixels to obtain the output, resulting in three output values after passing through the convolution layer.', 'ReLU layer serves as an activation function, activating a node if the input is above a certain quantity, outputting zero when the input is below zero, and having a linear relationship with the dependent variable when the input rises above the certain threshold.']}, {'end': 1086.467, 'segs': [{'end': 581.547, 'src': 'embed', 'start': 550.942, 'weight': 0, 'content': [{'end': 552.282, 'text': "So over here, I'll have 3.", 'start': 550.942, 'duration': 1.34}, {'end': 560.132, 'text': 'Again if I take my x value as 5 that obviously it is greater than or equal to 0 then my f of x becomes equal to x.', 'start': 552.282, 'duration': 7.85}, {'end': 562.255, 'text': 'So my f of x value becomes 5.', 'start': 560.132, 'duration': 2.123}, {'end': 563.777, 'text': 'So this is how a ReLU function works.', 'start': 562.255, 'duration': 1.522}, {'end': 570.606, 'text': 'So why are we using ReLU function here is we want to remove all the negative values from our output that we got through the convolution layer.', 'start': 564.017, 'duration': 6.589}, {'end': 575.364, 'text': "So we'll only take the first output that we got by moving one feature throughout the image.", 'start': 570.902, 'duration': 4.462}, {'end': 578.926, 'text': 'So this is the output that we have got for only one filter.', 'start': 575.724, 'duration': 3.202}, {'end': 581.547, 'text': "Alright, so over here I'm going to remove all negative values.", 'start': 579.306, 'duration': 2.241}], 'summary': 'Using relu function to remove negative values from output, resulting in f(x) = 5.', 'duration': 30.605, 'max_score': 550.942, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A550942.jpg'}, {'end': 640.814, 'src': 'heatmap', 'start': 599.916, 'weight': 0.724, 'content': [{'end': 604.418, 'text': 'So this is output only for one filter, after doing it for the output of the other two filters as well.', 'start': 599.916, 'duration': 4.502}, {'end': 605.958, 'text': 'We have got these two values more.', 'start': 604.778, 'duration': 1.18}, {'end': 609.56, 'text': 'So totally we have these three values after passing through ReLU activation function.', 'start': 605.998, 'duration': 3.562}, {'end': 611.961, 'text': "Next up we'll see what exactly is pooling layer.", 'start': 609.98, 'duration': 1.981}, {'end': 618.984, 'text': 'So in pooling layer what we do we take a window size of 2 and we move it across the entire matrix that we have got after passing through ReLU layer.', 'start': 612.281, 'duration': 6.703}, {'end': 624.426, 'text': 'and we take only the maximum value from there so that we can shrink the image.', 'start': 619.624, 'duration': 4.802}, {'end': 627.547, 'text': 'So what we are actually doing is we are reducing the size of our image.', 'start': 624.546, 'duration': 3.001}, {'end': 629.388, 'text': 'So let me explain you with an example.', 'start': 627.887, 'duration': 1.501}, {'end': 634.429, 'text': 'So this is basically one output that we have got after passing through ReLU layer and over here.', 'start': 629.908, 'duration': 4.521}, {'end': 636.35, 'text': 'we have taken a window size of two cross two.', 'start': 634.429, 'duration': 1.921}, {'end': 640.814, 'text': 'So when we keep this window at this particular position, we see that one is the highest value.', 'start': 636.73, 'duration': 4.084}], 'summary': 'After passing through relu layer, we have three values; pooling layer shrinks image size by taking maximum value from window of size 2x2.', 'duration': 40.898, 'max_score': 599.916, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A599916.jpg'}, {'end': 776.837, 'src': 'heatmap', 'start': 744.526, 'weight': 0.909, 'content': [{'end': 747.807, 'text': 'so therefore we have got three outputs after passing through pooling layer.', 'start': 744.526, 'duration': 3.281}, {'end': 749.228, 'text': 'All right, next up.', 'start': 748.147, 'duration': 1.081}, {'end': 750.849, 'text': 'We are going to stack up all the layers.', 'start': 749.268, 'duration': 1.581}, {'end': 752.189, 'text': "All right, so let's do that.", 'start': 750.909, 'duration': 1.28}, {'end': 756.591, 'text': 'So after passing through convolution Rayleigh and pooling we have got this 4 cross 4 matrix.', 'start': 752.569, 'duration': 4.022}, {'end': 757.632, 'text': 'This was our input image.', 'start': 756.611, 'duration': 1.021}, {'end': 765.114, 'text': 'Now when we add one more layer of convolution relu and pooling we have shrinked our image from 4 cross 4 to 2 cross 2 as you can notice here.', 'start': 758.112, 'duration': 7.002}, {'end': 767.254, 'text': 'Now we are going to use fully connected layer.', 'start': 765.474, 'duration': 1.78}, {'end': 770.995, 'text': 'Now what happens in fully connected layer the actual classification happens here guys, okay.', 'start': 767.294, 'duration': 3.701}, {'end': 776.837, 'text': 'So what we are doing here is we are going to take the shrinked images and put it into a single list.', 'start': 771.515, 'duration': 5.322}], 'summary': 'After passing through layers, image shrinks from 4x4 to 2x2. fully connected layer handles classification.', 'duration': 32.311, 'max_score': 744.526, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A744526.jpg'}, {'end': 786.219, 'src': 'embed', 'start': 756.611, 'weight': 1, 'content': [{'end': 757.632, 'text': 'This was our input image.', 'start': 756.611, 'duration': 1.021}, {'end': 765.114, 'text': 'Now when we add one more layer of convolution relu and pooling we have shrinked our image from 4 cross 4 to 2 cross 2 as you can notice here.', 'start': 758.112, 'duration': 7.002}, {'end': 767.254, 'text': 'Now we are going to use fully connected layer.', 'start': 765.474, 'duration': 1.78}, {'end': 770.995, 'text': 'Now what happens in fully connected layer the actual classification happens here guys, okay.', 'start': 767.294, 'duration': 3.701}, {'end': 776.837, 'text': 'So what we are doing here is we are going to take the shrinked images and put it into a single list.', 'start': 771.515, 'duration': 5.322}, {'end': 783.758, 'text': 'So basically this is what we have got after passing through two layers of convolution relu and pooling and this is what we have got.', 'start': 777.277, 'duration': 6.481}, {'end': 786.219, 'text': 'So basically we are converting into a single list or a vector.', 'start': 783.778, 'duration': 2.441}], 'summary': 'After two layers of convolution, relu, and pooling, the image shrinks from 4x4 to 2x2 before being converted into a single list for classification.', 'duration': 29.608, 'max_score': 756.611, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A756611.jpg'}, {'end': 989.055, 'src': 'embed', 'start': 954.484, 'weight': 2, 'content': [{'end': 955.625, 'text': 'So this is our use case, guys.', 'start': 954.484, 'duration': 1.141}, {'end': 962.611, 'text': 'So over here, what we are going to do is we are going to train our model on different types of dogs and cat images.', 'start': 955.645, 'duration': 6.966}, {'end': 969.638, 'text': 'And once the training is done, we are going to provide it an input and it will classify whether the input is of a dog or a cat.', 'start': 963.171, 'duration': 6.467}, {'end': 971.84, 'text': 'Now let me tell you the steps involved in it.', 'start': 970.018, 'duration': 1.822}, {'end': 976.145, 'text': 'So what we are going to do in the beginning is obviously first we need to download the data set.', 'start': 972.22, 'duration': 3.925}, {'end': 979.228, 'text': 'After that we are going to write a function to encode the labels.', 'start': 976.565, 'duration': 2.663}, {'end': 982.852, 'text': 'Labels are nothing but the dependent variable that we are trying to predict.', 'start': 979.348, 'duration': 3.504}, {'end': 989.055, 'text': 'So in our training data and testing data, obviously we know the labels, right? So on that basis only we can train our model.', 'start': 983.172, 'duration': 5.883}], 'summary': 'Training a model to classify dog or cat images after downloading and encoding the dataset.', 'duration': 34.571, 'max_score': 954.484, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A954484.jpg'}], 'start': 509.396, 'title': 'Relu function and cnn', 'summary': 'Explains the relu function, its application in removing negative values from the output of a convolution layer, and the process of passing an image through the convolution layer, relu layer, and pooling layer, resulting in a 91% match for classifying x over o. it also outlines the process of training a model to classify dog and cat images.', 'chapters': [{'end': 570.606, 'start': 509.396, 'title': 'Understanding relu function', 'summary': 'Explains the relu function, stating that when x is less than 0, f(x) is 0, and when x is greater than or equal to 0, f(x) is equal to x, and its application in removing negative values from the output of a convolution layer.', 'duration': 61.21, 'highlights': ['ReLU function states that f(x) is 0 when x is less than 0 and f(x) is equal to x when x is greater than or equal to 0. Clarifies the conditions for the ReLU function, defining its behavior with specific x values.', "Demonstrates the application of ReLU function with specific x values and explains its output for each case. Provides clear examples of the ReLU function's output for different x values, highlighting its behavior in processing negative and non-negative inputs.", 'Explains the purpose of using ReLU function to remove negative values from the output of a convolution layer. Highlights the practical application of ReLU in eliminating negative output values from a convolution layer.']}, {'end': 1086.467, 'start': 570.902, 'title': 'Understanding convolutional neural network', 'summary': 'Explains the process of passing an image through the convolution layer, relu layer, and pooling layer, reducing the image size from 7x7 to 4x4, and finally classifying the input image using a 12-element vector, resulting in a 91% match for classifying x over o. it also outlines the process of training a model to classify dog and cat images, including encoding labels, resizing images, building a deep neural network, and making predictions.', 'duration': 515.565, 'highlights': ['Process of Image Classification Explains the process of passing an image through the convolution layer, ReLU layer, and pooling layer, reducing the image size from 7x7 to 4x4, and classifying the input image using a 12-element vector, resulting in a 91% match for classifying X over O.', 'Training a Model for Image Classification Details the process of training a model to classify dog and cat images, including encoding labels, resizing images, building a deep neural network, and making predictions.']}], 'duration': 577.071, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A509396.jpg', 'highlights': ['Explains the purpose of using ReLU function to remove negative values from the output of a convolution layer. Highlights the practical application of ReLU in eliminating negative output values from a convolution layer.', 'Process of Image Classification Explains the process of passing an image through the convolution layer, ReLU layer, and pooling layer, reducing the image size from 7x7 to 4x4, and classifying the input image using a 12-element vector, resulting in a 91% match for classifying X over O.', 'Training a Model for Image Classification Details the process of training a model to classify dog and cat images, including encoding labels, resizing images, building a deep neural network, and making predictions.', "Demonstrates the application of ReLU function with specific x values and explains its output for each case. Provides clear examples of the ReLU function's output for different x values, highlighting its behavior in processing negative and non-negative inputs.", 'ReLU function states that f(x) is 0 when x is less than 0 and f(x) is equal to x when x is greater than or equal to 0. Clarifies the conditions for the ReLU function, defining its behavior with specific x values.']}, {'end': 1315.122, 'segs': [{'end': 1135.583, 'src': 'embed', 'start': 1106.574, 'weight': 1, 'content': [{'end': 1111.596, 'text': 'So, basically, what we have done here is we have resized our image to 50 cross, 50 cross 1 matrix,', 'start': 1106.574, 'duration': 5.022}, {'end': 1113.896, 'text': 'and that is the size of the input that we are using.', 'start': 1111.596, 'duration': 2.3}, {'end': 1115.797, 'text': "right, this input that I'm talking about.", 'start': 1113.896, 'duration': 1.901}, {'end': 1117.978, 'text': 'then what we have done a convolution layer.', 'start': 1115.797, 'duration': 2.181}, {'end': 1119.819, 'text': 'We have defined with 32 filters.', 'start': 1118.018, 'duration': 1.801}, {'end': 1127.401, 'text': 'and a stride of five with an activation function a relu and after that we have added a pooling layer max pool layer.', 'start': 1121.039, 'duration': 6.362}, {'end': 1129.701, 'text': 'Okay, again what we have done.', 'start': 1127.701, 'duration': 2}, {'end': 1135.583, 'text': 'we have repeated the same process, but over here we are taking 64 filters and five stride,', 'start': 1129.701, 'duration': 5.882}], 'summary': 'Resized image to 50x50x1, used 32 filters & 5 stride in convolution layer, added max pool layer, repeated with 64 filters & 5 stride.', 'duration': 29.009, 'max_score': 1106.574, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A1106574.jpg'}, {'end': 1170.28, 'src': 'embed', 'start': 1146.707, 'weight': 2, 'content': [{'end': 1154.931, 'text': 'and after that we are using a fully connected layer with 1024 neurons and finally we are using the dropout layer, with key probability of 0.8,', 'start': 1146.707, 'duration': 8.224}, {'end': 1155.772, 'text': 'to finish our models.', 'start': 1154.931, 'duration': 0.841}, {'end': 1164.336, 'text': 'This is where our model is actually finished and then what we are doing is we are using the atom optimizer to optimize our model.', 'start': 1155.892, 'duration': 8.444}, {'end': 1170.28, 'text': 'So basically whatever the loss that we have we are trying to reduce it and this is basically for your TensorBoard.', 'start': 1164.356, 'duration': 5.924}], 'summary': 'Using fully connected layer with 1024 neurons, dropout layer with 0.8 probability, and atom optimizer to optimize model for reducing loss in tensorboard.', 'duration': 23.573, 'max_score': 1146.707, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A1146707.jpg'}, {'end': 1256.945, 'src': 'embed', 'start': 1220.441, 'weight': 0, 'content': [{'end': 1226.003, 'text': "Alright, so I've already trained the model because it takes a lot of time and yeah, I cannot do it here.", 'start': 1220.441, 'duration': 5.562}, {'end': 1236.207, 'text': "So I've already trained the model and you can see that the loss that came after the 10th epoch is 0.2973 and the accuracy somewhere around 88%,", 'start': 1226.063, 'duration': 10.144}, {'end': 1237.287, 'text': 'which is pretty good guys.', 'start': 1236.207, 'duration': 1.08}, {'end': 1240.829, 'text': "And yeah, and I've done the prediction on the test data as well.", 'start': 1237.828, 'duration': 3.001}, {'end': 1246.111, 'text': 'So let me just show it to you that so this is the prediction that it has done on few of the images in the test data.', 'start': 1240.889, 'duration': 5.222}, {'end': 1251.499, 'text': 'So yeah, it is a cat predicted as cat, cat predicted as cat, cat, cat, cat and dogs as well.', 'start': 1246.772, 'duration': 4.727}, {'end': 1252.781, 'text': 'There are certain dogs as well.', 'start': 1251.519, 'duration': 1.262}, {'end': 1256.945, 'text': "So this is it for today's session, guys, and if you want the code and the data set,", 'start': 1253.342, 'duration': 3.603}], 'summary': 'Model trained with 0.2973 loss, 88% accuracy; made successful predictions on test data.', 'duration': 36.504, 'max_score': 1220.441, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A1220441.jpg'}, {'end': 1301.008, 'src': 'embed', 'start': 1274.502, 'weight': 4, 'content': [{'end': 1278.326, 'text': 'Then we saw why we need convolution neural networks and what is convolution neural network.', 'start': 1274.502, 'duration': 3.824}, {'end': 1281.429, 'text': 'After that, we understood how a convolutional neural network works.', 'start': 1278.806, 'duration': 2.623}, {'end': 1286.574, 'text': 'We discussed all the layers that are involved in it, like convolution, ray loop, pooling, fully connected layer.', 'start': 1281.469, 'duration': 5.105}, {'end': 1288.296, 'text': 'all those layers we have discussed in detail.', 'start': 1286.574, 'duration': 1.722}, {'end': 1292.48, 'text': 'And then finally, we implemented a use case which can classify the images of dogs and cats.', 'start': 1288.356, 'duration': 4.124}, {'end': 1294.862, 'text': "All right, so this is it for today's session, guys.", 'start': 1293.08, 'duration': 1.782}, {'end': 1298.986, 'text': 'If you have any questions, any doubts, just go ahead and type it in the comment section below.', 'start': 1294.982, 'duration': 4.004}, {'end': 1301.008, 'text': "You'll get the answer as soon as possible.", 'start': 1299.327, 'duration': 1.681}], 'summary': 'Covered cnn layers and implemented image classification for dogs and cats.', 'duration': 26.506, 'max_score': 1274.502, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A1274502.jpg'}], 'start': 1086.567, 'title': 'Convolutional neural networks', 'summary': 'Discusses building a convolutional neural network model with layers of different filter sizes (128, 64, and 32) and strides, with the input image resized to a 50x50x1 matrix. it also details the training of a neural network model with 1024 neurons, using dropout with a probability of 0.8, and an atom optimizer, achieving a 88% accuracy and a loss of 0.2973 after 10 epochs, with predictions on test data. additionally, it covers the basics of convolutional neural networks, including the limitations of fully connected networks for image recognition, the need for and functioning of convolutional neural networks, and a practical use case for classifying images of dogs and cats.', 'chapters': [{'end': 1146.707, 'start': 1086.567, 'title': 'Building convolutional neural network model', 'summary': 'Discusses building a convolutional neural network model with layers of different filter sizes (128, 64, and 32) and strides, with the input image resized to a 50x50x1 matrix.', 'duration': 60.14, 'highlights': ['The input image is resized to a 50x50x1 matrix, and the model includes convolution layers with 128, 64, and 32 filters, each with a stride of five and a max pool layer after each convolution layer.', 'The data set is split into two parts for training and testing purposes, and the model is defined for further processing.']}, {'end': 1256.945, 'start': 1146.707, 'title': 'Neural network model training and testing', 'summary': 'Details the training of a neural network model with 1024 neurons, using dropout with a probability of 0.8, and an atom optimizer, achieving a 88% accuracy and a loss of 0.2973 after 10 epochs, with predictions on test data.', 'duration': 110.238, 'highlights': ['The model utilizes a fully connected layer with 1024 neurons and a dropout layer with a probability of 0.8. Specifies the architecture of the neural network model, including the number of neurons and the use of dropout for regularization.', 'The model achieves an accuracy of approximately 88% and a loss of 0.2973 after 10 epochs of training. Quantifies the performance of the model in terms of accuracy and loss after training for 10 epochs.', 'The predictions on the test data show accurate identification of cats and dogs by the model. Demonstrates the successful predictions made by the trained model on test data, highlighting its effectiveness.']}, {'end': 1315.122, 'start': 1256.945, 'title': 'Convolutional neural networks', 'summary': 'Discusses the basics of convolutional neural networks, covering how a computer reads an image, the limitations of fully connected networks for image recognition, the need for and functioning of convolutional neural networks, and a practical use case for classifying images of dogs and cats.', 'duration': 58.177, 'highlights': ['We implemented a use case which can classify the images of dogs and cats.', 'We discussed all the layers involved in convolutional neural networks, such as convolution, ray loop, pooling, and fully connected layer.', 'We saw why we need convolution neural networks and what is convolution neural network.', 'We started by understanding how a computer reads an image.']}], 'duration': 228.555, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/umGJ30-15_A/pics/umGJ30-15_A1086567.jpg', 'highlights': ['The model achieves an accuracy of approximately 88% and a loss of 0.2973 after 10 epochs of training.', 'The input image is resized to a 50x50x1 matrix, and the model includes convolution layers with 128, 64, and 32 filters, each with a stride of five and a max pool layer after each convolution layer.', 'The model utilizes a fully connected layer with 1024 neurons and a dropout layer with a probability of 0.8.', 'The predictions on the test data show accurate identification of cats and dogs by the model.', 'We discussed all the layers involved in convolutional neural networks, such as convolution, ray loop, pooling, and fully connected layer.']}], 'highlights': ['The model achieves an accuracy of approximately 88% and a loss of 0.2973 after 10 epochs of training.', 'Process of Image Classification Explains the process of passing an image through the convolution layer, ReLU layer, and pooling layer, reducing the image size from 7x7 to 4x4, and classifying the input image using a 12-element vector, resulting in a 91% match for classifying X over O.', 'The chapter introduces the concept of convolutional neural networks and explains its three layers: convolutional layer, ReLU layer, pooling layer, and fully connected layer.', 'Explains the limitations of fully connected networks for image classification, citing the exponential increase in weights for larger images and the resulting overfitting issues.', 'The input image is resized to a 50x50x1 matrix, and the model includes convolution layers with 128, 64, and 32 filters, each with a stride of five and a max pool layer after each convolution layer.']}