{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Convolutional neural networks\n", "\n", "In the previous unit we have learned how to define a multi-layered neural network using class definition, but those networks were generic, and not specialized for computer vision tasks. In this unit we will learn about **Convolutional Neural Networks (CNNs)**, which are specifically designed for computer vision.\n", "\n", "Computer vision is different from generic classification, because when we are trying to find a certain object in the picture, we are scanning the image looking for some specific **patterns** and their combinations. For example, when looking for a cat, we first may look for horizontal lines, which can form whiskers, and then certain combination of whiskers can tell us that it is actually a picture of a cat. Relative position and presence of certain patterns is important, and not their exact position on the image. \n", "\n", "To extract patterns, we will use the notion of **convolutional filters**. But first, let us load all dependencies and functions that we have defined in the previous units." ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "import torch\n", "import torch.nn as nn\n", "import torchvision\n", "import matplotlib.pyplot as plt\n", "from torchinfo import summary\n", "import numpy as np\n", "\n", "from pytorchcv import load_mnist, train, plot_results, plot_convolution, display_dataset\n", "load_mnist(batch_size=128)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Convolutional filters\n", "\n", "Convolutional filters are small windows that run over each pixel of the image and compute weighted average of the neighboring pixels.\n", "\n", "\n", "\n", "They are defined by matrices of weight coefficients. Let's see the examples of applying two different convolutional filters over our MNIST handwritten digits:" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "<Figure size 576x216 with 12 Axes>" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "<Figure size 576x216 with 12 Axes>" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "plot_convolution(torch.tensor([[-1.,0.,1.],[-1.,0.,1.],[-1.,0.,1.]]),'Vertical edge filter')\n", "plot_convolution(torch.tensor([[-1.,-1.,-1.],[0.,0.,0.],[1.,1.,1.]]),'Horizontal edge filter')\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "First filter is called a **vertical edge filter**, and it is defined by the following matrix:\n", "$$\n", "\\left(\n", " \\begin{matrix}\n", " -1 & 0 & 1 \\cr\n", " -1 & 0 & 1 \\cr\n", " -1 & 0 & 1 \\cr\n", " \\end{matrix}\n", "\\right)\n", "$$\n", "When this filter goes over relatively uniform pixel field, all values add up to 0. However, when it encounters a vertical edge in the image, high spike value is generated. That's why in the images above you can see vertical edges represented by high and low values, while horizontal edges are averaged out.\n", "\n", "An opposite thing happens when we apply horizontal edge filter - horizontal lines are amplified, and vertical are averaged out.\n", "\n", "In classical computer vision, multiple filters were applied to the image to generate features, which then were used by machine learning algorithm to build a classifier. However, in deep learning we construct networks that **learn** best convolutional filters to solve classification problem.\n", "\n", "To do that, we introduce **convolutional layers**." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Covolutional layers\n", "\n", "Convolutional layers are defined using `nn.Conv2d` construction. We need to specify the following:\n", "* `in_channels` - number of input channels. In our case we are dealing with a grayscale image, thus number of input channels is 1.\n", "* `out_channels` - number of filters to use. We will use 9 different filters, which will give the network plenty of opportunities to explore which filters work best for our scenario.\n", "* `kernel_size` is the size of the sliding window. Usually 3x3 or 5x5 filters are used.\n", "\n", "Simplest CNN will contain one convolutional layer. Given the input size 28x28, after applying nine 5x5 filters we will end up with a tensor of 9x24x24 (the spatial size is smaller, because there are only 24 positions where a sliding interval of length 5 can fit into 28 pixels).\n", "\n", "After convolution, we flatten 9x24x24 tensor into one vector of size 5184, and then add linear layer, to produce 10 classes. We also use `relu` activation function in between layers. " ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "==========================================================================================\n", "Layer (type:depth-idx) Output Shape Param #\n", "==========================================================================================\n", "├─Conv2d: 1-1 [1, 9, 24, 24] 234\n", "├─Flatten: 1-2 [1, 5184] --\n", "├─Linear: 1-3 [1, 10] 51,850\n", "==========================================================================================\n", "Total params: 52,084\n", "Trainable params: 52,084\n", "Non-trainable params: 0\n", "Total mult-adds (M): 0.18\n", "==========================================================================================\n", "Input size (MB): 0.00\n", "Forward/backward pass size (MB): 0.04\n", "Params size (MB): 0.21\n", "Estimated Total Size (MB): 0.25\n", "==========================================================================================" ] }, "execution_count": 6, "metadata": {}, "output_type": "execute_result" } ], "source": [ "class OneConv(nn.Module):\n", " def __init__(self):\n", " super(OneConv, self).__init__()\n", " self.conv = nn.Conv2d(in_channels=1,out_channels=9,kernel_size=(5,5))\n", " self.flatten = nn.Flatten()\n", " self.fc = nn.Linear(5184,10)\n", "\n", " def forward(self, x):\n", " x = nn.functional.relu(self.conv(x))\n", " x = self.flatten(x)\n", " x = nn.functional.log_softmax(self.fc(x),dim=1)\n", " return x\n", "\n", "net = OneConv()\n", "\n", "summary(net,input_size=(1,1,28,28))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "You can see that this network contains around 50k trainable parameters, compared to around 80k in fully-connected multi-layered networks. This allows us to achieve good results even on smaller datasets, because convolutional networks generalize much better." ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Epoch 0, Train acc=0.947, Val acc=0.969, Train loss=0.001, Val loss=0.001\n", "Epoch 1, Train acc=0.979, Val acc=0.975, Train loss=0.001, Val loss=0.001\n", "Epoch 2, Train acc=0.985, Val acc=0.977, Train loss=0.000, Val loss=0.001\n", "Epoch 3, Train acc=0.988, Val acc=0.975, Train loss=0.000, Val loss=0.001\n", "Epoch 4, Train acc=0.988, Val acc=0.976, Train loss=0.000, Val loss=0.001\n" ] }, { "data": { "image/png": "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\n", "text/plain": [ "<Figure size 1080x360 with 2 Axes>" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "hist = train(net,train_loader,test_loader,epochs=5)\n", "plot_results(hist)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "As you can see, we are able to achieve higher accuracy, and much faster, compared to the fully-connected networks from previous unit.\n", "\n", "We can also visualize the weights of our trained convolutional layers, to try and make some more sense of what is going on:" ] }, { "cell_type": "code", "execution_count": 9, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "<Figure size 432x288 with 9 Axes>" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "fig,ax = plt.subplots(1,9)\n", "with torch.no_grad():\n", " p = next(net.conv.parameters())\n", " for i,x in enumerate(p):\n", " ax[i].imshow(x.detach().cpu()[0,...])\n", " ax[i].axis('off')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "You can see that some of those filters look like they can recognize some oblique strokes, while others look pretty random. \n", "\n", "## Multi-layered CNNs and pooling layers\n", "\n", "First convolutional layers looks for primitive patterns, such as horizontal or vertical lines, but we can apply further convolutional layers on top of them to look for higher-level patterns, such as primitive shapes. Then more convolutional layers can combine those shapes into some parts of the picture, up to the final object that we are trying to classify. \n", "\n", "When doing so, we may also apply one trick: reducing the spatial size of the image. Once we have detected there is a horizontal stoke within sliding 3x3 window, it is not so important at which exact pixel it occurred. Thus we can \"scale down\" the size of the image, which is done using one of the **pooling layers**:\n", "\n", " * **Average Pooling** takes a sliding window (for example, 2x2 pixels) and computes an average of values within the window\n", " * **Max Pooling** replaces the window with the maximum value. The idea behind max pooling is to detect a presence of a certain pattern within the sliding window.\n", "\n", "Thus, in a typical CNN there would be several convolutional layers, with pooling layers in between them to decrease dimensions of the image. We would also increase the number of filters, because as patterns become more advanced - there are more possible interesting combinations that we need to be looking for.\n", "\n", "\n", "\n", "Because of decreasing spatial dimensions and increasing feature/filters dimensions, this architecture is also called **pyramid architecture**. " ] }, { "cell_type": "code", "execution_count": 13, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "==========================================================================================\n", "Layer (type:depth-idx) Output Shape Param #\n", "==========================================================================================\n", "├─Conv2d: 1-1 [1, 10, 24, 24] 260\n", "├─MaxPool2d: 1-2 [1, 10, 12, 12] --\n", "├─Conv2d: 1-3 [1, 20, 8, 8] 5,020\n", "├─MaxPool2d: 1-4 [1, 20, 4, 4] --\n", "├─Linear: 1-5 [1, 10] 3,210\n", "==========================================================================================\n", "Total params: 8,490\n", "Trainable params: 8,490\n", "Non-trainable params: 0\n", "Total mult-adds (M): 0.47\n", "==========================================================================================\n", "Input size (MB): 0.00\n", "Forward/backward pass size (MB): 0.06\n", "Params size (MB): 0.03\n", "Estimated Total Size (MB): 0.09\n", "==========================================================================================" ] }, "execution_count": 13, "metadata": {}, "output_type": "execute_result" } ], "source": [ "class MultiLayerCNN(nn.Module):\n", " def __init__(self):\n", " super(MultiLayerCNN, self).__init__()\n", " self.conv1 = nn.Conv2d(1, 10, 5)\n", " self.pool = nn.MaxPool2d(2, 2)\n", " self.conv2 = nn.Conv2d(10, 20, 5)\n", " self.fc = nn.Linear(320,10)\n", "\n", " def forward(self, x):\n", " x = self.pool(nn.functional.relu(self.conv1(x)))\n", " x = self.pool(nn.functional.relu(self.conv2(x)))\n", " x = x.view(-1, 320)\n", " x = nn.functional.log_softmax(self.fc(x),dim=1)\n", " return x\n", "\n", "net = MultiLayerCNN()\n", "summary(net,input_size=(1,1,28,28))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Note a few things about this definition:\n", "* Instead of using `Flatten` layer, we are flattening the tensor inside `forward` function using `view` function. Since flattening layer does not have trainable weights, it is not essential that we create a separate layer instance within our class\n", "* We use just one instance of pooling layer in our model, also because it does not contain any trainable parameters, and this one instance can be effectively reused\n", "* The number of trainable parameters (~8.5K) is dramatically smaller than in previous cases. This happens because convolutional layers in general have few parameters, and dimensionality of the image before applying final dense layer is significantly reduced. Small number of parameters have positive impact on our models, because it helps to prevent overfitting even on smaller dataset sizes." ] }, { "cell_type": "code", "execution_count": 14, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Epoch 0, Train acc=0.952, Val acc=0.977, Train loss=0.001, Val loss=0.001\n", "Epoch 1, Train acc=0.982, Val acc=0.983, Train loss=0.000, Val loss=0.000\n", "Epoch 2, Train acc=0.986, Val acc=0.983, Train loss=0.000, Val loss=0.000\n", "Epoch 3, Train acc=0.986, Val acc=0.978, Train loss=0.000, Val loss=0.001\n", "Epoch 4, Train acc=0.987, Val acc=0.981, Train loss=0.000, Val loss=0.000\n" ] } ], "source": [ "hist = train(net,train_loader,test_loader,epochs=5)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "What you should probably observe is that we are able to achieve higher accuracy than with just one layer, and much faster - just with 1 or 2 epochs. It means that sophisticated network architecture needs much fewer data to figure out what is going on, and to extract generic patterns from our images.\n", "\n", "## Playing with real images from the CIFAR-10 dataset\n", "\n", "While our handwritten digit recognition problem may seem like a toy problem, we are now ready to do something more serious. Let's explore more advanced dataset of pictures of different objects, called [CIFAR-10](https://www.cs.toronto.edu/~kriz/cifar.html). It contains 60k 32x32 images, divided into 10 classes. " ] }, { "cell_type": "code", "execution_count": 22, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Downloading https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz to ./data/cifar-10-python.tar.gz\n" ] }, { "data": { "application/vnd.jupyter.widget-view+json": { "model_id": "77b339f50a6b48e98e5db22e0ddd7eb6", "version_major": 2, "version_minor": 0 }, "text/plain": [ "HBox(children=(FloatProgress(value=1.0, bar_style='info', max=1.0), HTML(value='')))" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ "Extracting ./data/cifar-10-python.tar.gz to ./data\n", "Files already downloaded and verified\n" ] } ], "source": [ "transform = torchvision.transforms.Compose(\n", " [torchvision.transforms.ToTensor(),\n", " torchvision.transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))])\n", "\n", "trainset = torchvision.datasets.CIFAR10(root='./data', train=True, download=True, transform=transform)\n", "trainloader = torch.utils.data.DataLoader(trainset, batch_size=14, shuffle=True)\n", "testset = torchvision.datasets.CIFAR10(root='./data', train=False, download=True, transform=transform)\n", "testloader = torch.utils.data.DataLoader(testset, batch_size=14, shuffle=False)\n", "classes = ('plane', 'car', 'bird', 'cat',\n", " 'deer', 'dog', 'frog', 'horse', 'ship', 'truck')" ] }, { "cell_type": "code", "execution_count": 23, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "<Figure size 1080x216 with 10 Axes>" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "display_dataset(trainset,classes=classes)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "A well-known architecture for CIFAR-10 is called [LeNet](https://en.wikipedia.org/wiki/LeNet), and has been proposed by *Yann LeCun*. It follows the same principles as we have outlined above, the main difference being 3 input color channels instead of 1. \n", "\n", "We also do one more simplification to this model - we do not use `log_softmax` as output activation function, and just return the output of last fully-connected layer. In this case we can just use `CrossEntropyLoss` loss function to optimize the model." ] }, { "cell_type": "code", "execution_count": 25, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "==========================================================================================\n", "Layer (type:depth-idx) Output Shape Param #\n", "==========================================================================================\n", "├─Conv2d: 1-1 [1, 6, 28, 28] 456\n", "├─MaxPool2d: 1-2 [1, 6, 14, 14] --\n", "├─Conv2d: 1-3 [1, 16, 10, 10] 2,416\n", "├─MaxPool2d: 1-4 [1, 16, 5, 5] --\n", "├─Conv2d: 1-5 [1, 120, 1, 1] 48,120\n", "├─Flatten: 1-6 [1, 120] --\n", "├─Linear: 1-7 [1, 64] 7,744\n", "├─Linear: 1-8 [1, 10] 650\n", "==========================================================================================\n", "Total params: 59,386\n", "Trainable params: 59,386\n", "Non-trainable params: 0\n", "Total mult-adds (M): 0.65\n", "==========================================================================================\n", "Input size (MB): 0.01\n", "Forward/backward pass size (MB): 0.05\n", "Params size (MB): 0.24\n", "Estimated Total Size (MB): 0.30\n", "==========================================================================================" ] }, "execution_count": 25, "metadata": {}, "output_type": "execute_result" } ], "source": [ "class LeNet(nn.Module):\n", " def __init__(self):\n", " super(LeNet, self).__init__()\n", " self.conv1 = nn.Conv2d(3, 6, 5)\n", " self.pool = nn.MaxPool2d(2)\n", " self.conv2 = nn.Conv2d(6, 16, 5)\n", " self.conv3 = nn.Conv2d(16,120,5)\n", " self.flat = nn.Flatten()\n", " self.fc1 = nn.Linear(120,64)\n", " self.fc2 = nn.Linear(64,10)\n", "\n", " def forward(self, x):\n", " x = self.pool(nn.functional.relu(self.conv1(x)))\n", " x = self.pool(nn.functional.relu(self.conv2(x)))\n", " x = nn.functional.relu(self.conv3(x))\n", " x = self.flat(x)\n", " x = nn.functional.relu(self.fc1(x))\n", " x = self.fc2(x)\n", " return x\n", "\n", "net = LeNet()\n", "\n", "summary(net,input_size=(1,3,32,32))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Training this network properly will take significant amount of time, and should preferably be done on GPU-enabled compute." ] }, { "cell_type": "code", "execution_count": 26, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Epoch 0, Train acc=0.261, Val acc=0.388, Train loss=0.143, Val loss=0.121\n", "Epoch 1, Train acc=0.437, Val acc=0.491, Train loss=0.110, Val loss=0.101\n", "Epoch 2, Train acc=0.508, Val acc=0.522, Train loss=0.097, Val loss=0.094\n" ] } ], "source": [ "opt = torch.optim.SGD(net.parameters(),lr=0.001,momentum=0.9)\n", "hist = train(net, trainloader, testloader, epochs=3, optimizer=opt, loss_fn=nn.CrossEntropyLoss())" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The accuracy that we have been able to achieve with 3 epochs of training does not seem great. However, remember that blind guessing would only give us 10% accuracy, and that our problem is actually significantly more difficult than MNIST digit classification. Getting above 50% accuracy in such a short training time seems like a good accomplishment.\n", "\n", "## Takeaways\n", "\n", "In this unit, we have learned the main concept behind computer vision neural networks - convolutional networks. Real-life architectures that power image classification, object detection, and even image generation networks are all based on CNNs, just with more layers and some additional training tricks." ] } ], "metadata": { "kernelspec": { "display_name": "py37_pytorch", "language": "python", "name": "conda-env-py37_pytorch-py" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.7.7" } }, "nbformat": 4, "nbformat_minor": 2 }