{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Building a Feed-forward Neural Network using PyTorch \n", "Adapted from: https://pytorch.org/tutorials/beginner/blitz/neural_networks_tutorial.html & https://pytorch.org/tutorials/beginner/blitz/cifar10_tutorial.html\n", "\n", "---\n", "\n", "## Training an image classifier to classify happy and sad faces\n", "\n", "We will perform the following steps in order:\n", "\n", "1. Load and normalizing the the ![NimStim Dataset of Facial Expressions](https://danlab7.wixsite.com/nimstim) using torchvision\n", "2. Define a Convolutional Neural Network\n", "3. Define a loss function\n", "4. Train the network on the training data\n", "5. Test the network on the test data" ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "from __future__ import print_function, division\n", "\n", "import torch\n", "import torch.nn as nn\n", "import torch.optim as optim\n", "from torch.optim import lr_scheduler\n", "\n", "import torchvision\n", "from torchvision import datasets, models, transforms\n", "\n", "import numpy as np\n", "import matplotlib.pyplot as plt\n", "import os\n", "\n", "plt.ion()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 1. Load and normalize NimStim Dataset of Facial Expressions\n", "\n", "We will load our dataset using `torchvision` from the following file structure:\n", "```\n", "├── NN_Emotional_Faces_Classifier.ipynb\n", "├── nim_stim\n", "| └── training\n", "| └── happiness\n", "| └── sadness\n", "| └── testing\n", "| └── happiness\n", "| └── sadness\n", "```\n", "The output of torchvision datasets are PILImage images of range [0, 1]. We transform them to Tensors of normalized range [-1, 1]." ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [], "source": [ "data_transforms = {\n", " 'training': transforms.Compose([\n", " transforms.RandomHorizontalFlip(),\n", " transforms.Resize((128,128)),\n", " transforms.ToTensor(),\n", " transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))\n", " ]),\n", " 'testing': transforms.Compose([\n", " transforms.Resize((128,128)),\n", " transforms.ToTensor(),\n", " transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))\n", " ]),\n", "}\n", "\n", "data_dir = './nim_stim/'\n", "image_datasets = {x: datasets.ImageFolder(os.path.join(data_dir, x),\n", " data_transforms[x])\n", " for x in ['training', 'testing']}\n", "\n", "dataloaders = {x: torch.utils.data.DataLoader(image_datasets[x], batch_size=4,\n", " shuffle=True, num_workers=4)\n", " for x in ['training', 'testing']}\n", "\n", "dataset_sizes = {x: len(image_datasets[x]) for x in ['training', 'testing']}\n", "class_names = image_datasets['training'].classes\n", "\n", "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Let's view some of our images:" ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "def imshow(inp, title=None):\n", " \"\"\"Imshow for Tensor.\"\"\"\n", " inp = inp.numpy().transpose((1, 2, 0))\n", " mean = np.array([0.5, 0.5, 0.5])\n", " std = np.array([0.5, 0.5, 0.5])\n", " inp = std * inp + mean\n", " inp = np.clip(inp, 0, 1)\n", " plt.imshow(inp)\n", " if title is not None:\n", " plt.title(title)\n", " plt.pause(0.001) # pause a bit so that plots are updated\n", "\n", "# Get a batch of training data\n", "inputs, classes = next(iter(dataloaders['training']))\n", "\n", "# Make a grid from batch\n", "out = torchvision.utils.make_grid(inputs)\n", "\n", "imshow(out, title=[class_names[x] for x in classes])" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 2. Define a Convolutional Neural Network\n", "\n", "A typical training procedure for a neural network is as follows:\n", "1. Define the neural network that has some learnable parameters (or weights)\n", "2. Iterate over a dataset of inputs\n", "3. Process input through the network\n", "4. Compute the loss (how far is the output from being correct)\n", "5. Propagate gradients back into the network’s parameters\n", "6. Update the weights of the network, typically using a simple update rule: weight = weight - learning_rate * gradient\n", "\n", "Luckily, PyTorch let's us implement this fairly easily" ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [], "source": [ "#define the network\n", "class Net(nn.Module):\n", " def __init__(self):\n", " super(Net, self).__init__()\n", " self.conv1 = nn.Conv2d(3, 6, 5)\n", " self.pool = nn.MaxPool2d(2, 2)\n", " self.conv2 = nn.Conv2d(6, 16, 5)\n", " self.fc1 = nn.Linear(16*841, 120)\n", " self.fc2 = nn.Linear(120, 84)\n", " self.fc3 = nn.Linear(84, 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(x.size(0), 16*841)\n", " x = nn.functional.relu(self.fc1(x))\n", " x = nn.functional.relu(self.fc2(x))\n", " x = self.fc3(x)\n", " return x\n", "\n", "net = Net()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 3. Define a Loss function and optimizer\n", "\n", "Now that we've defined the network with a feed-forward function, let's specify the **loss function** to take the (output, target) pair of inputs, and compute a value that estimates how far away the output is from the target.\n", "\n", "To backpropagate the error all we have to do is call `loss.backward()`. You need to clear the existing gradients though with `optimizer.zero_grad()`, or else gradients will be accumulated to existing gradients. When we call `loss.backward()`, the whole graph is differentiated w.r.t. the loss.\n", "\n", "The package `torch.optim` implements all sorts of different update rules. For now, we will implement the simplest: Stochastic Gradient Descent (SGD): `weight = weight - learning_rate * gradient` using `optim.SGD()`." ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [], "source": [ "criterion = nn.CrossEntropyLoss() #nn.MSELoss()\n", "optimizer = optim.SGD(net.parameters(), lr=0.001, momentum=0.9)\n", "num_epochs = 75" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 4. Train the network (where the magic happens)\n", "\n", "We will loop over our data iterator, and feed the inputs to the network and optimize using our specified settings" ] }, { "cell_type": "code", "execution_count": 6, "metadata": { "scrolled": true }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Epoch 0/74\n", "----------\n", "[1] loss: 0.129\n", "Epoch 1/74\n", "----------\n", "[2] loss: 0.112\n", "Epoch 2/74\n", "----------\n", "[3] loss: 0.048\n", "Epoch 3/74\n", "----------\n", "[4] loss: 0.032\n", "Epoch 4/74\n", "----------\n", "[5] loss: 0.031\n", "Epoch 5/74\n", "----------\n", "[6] loss: 0.072\n", "Epoch 6/74\n", "----------\n", "[7] loss: 0.040\n", "Epoch 7/74\n", "----------\n", "[8] loss: 0.032\n", "Epoch 8/74\n", "----------\n", "[9] loss: 0.049\n", "Epoch 9/74\n", "----------\n", "[10] loss: 0.040\n", "Epoch 10/74\n", "----------\n", "[11] loss: 0.035\n", "Epoch 11/74\n", "----------\n", "[12] loss: 0.033\n", "Epoch 12/74\n", "----------\n", "[13] loss: 0.038\n", "Epoch 13/74\n", "----------\n", "[14] loss: 0.039\n", "Epoch 14/74\n", "----------\n", "[15] loss: 0.035\n", "Epoch 15/74\n", "----------\n", "[16] loss: 0.037\n", "Epoch 16/74\n", "----------\n", "[17] loss: 0.033\n", "Epoch 17/74\n", "----------\n", "[18] loss: 0.076\n", "Epoch 18/74\n", "----------\n", "[19] loss: 0.036\n", "Epoch 19/74\n", "----------\n", "[20] loss: 0.038\n", "Epoch 20/74\n", "----------\n", "[21] loss: 0.033\n", "Epoch 21/74\n", "----------\n", "[22] loss: 0.052\n", "Epoch 22/74\n", "----------\n", "[23] loss: 0.030\n", "Epoch 23/74\n", "----------\n", "[24] loss: 0.039\n", "Epoch 24/74\n", "----------\n", "[25] loss: 0.050\n", "Epoch 25/74\n", "----------\n", "[26] loss: 0.033\n", "Epoch 26/74\n", "----------\n", "[27] loss: 0.040\n", "Epoch 27/74\n", "----------\n", "[28] loss: 0.045\n", "Epoch 28/74\n", "----------\n", "[29] loss: 0.043\n", "Epoch 29/74\n", "----------\n", "[30] loss: 0.035\n", "Epoch 30/74\n", "----------\n", "[31] loss: 0.022\n", "Epoch 31/74\n", "----------\n", "[32] loss: 0.022\n", "Epoch 32/74\n", "----------\n", "[33] loss: 0.011\n", "Epoch 33/74\n", "----------\n", "[34] loss: 0.019\n", "Epoch 34/74\n", "----------\n", "[35] loss: 0.025\n", "Epoch 35/74\n", "----------\n", "[36] loss: 0.024\n", "Epoch 36/74\n", "----------\n", "[37] loss: 0.005\n", "Epoch 37/74\n", "----------\n", "[38] loss: 0.007\n", "Epoch 38/74\n", "----------\n", "[39] loss: 0.012\n", "Epoch 39/74\n", "----------\n", "[40] loss: 0.005\n", "Epoch 40/74\n", "----------\n", "[41] loss: 0.033\n", "Epoch 41/74\n", "----------\n", "[42] loss: 0.015\n", "Epoch 42/74\n", "----------\n", "[43] loss: 0.009\n", "Epoch 43/74\n", "----------\n", "[44] loss: 0.001\n", "Epoch 44/74\n", "----------\n", "[45] loss: 0.001\n", "Epoch 45/74\n", "----------\n", "[46] loss: 0.027\n", "Epoch 46/74\n", "----------\n", "[47] loss: 0.014\n", "Epoch 47/74\n", "----------\n", "[48] loss: 0.000\n", "Epoch 48/74\n", "----------\n", "[49] loss: 0.003\n", "Epoch 49/74\n", "----------\n", "[50] loss: 0.016\n", "Epoch 50/74\n", "----------\n", "[51] loss: 0.011\n", "Epoch 51/74\n", "----------\n", "[52] loss: 0.017\n", "Epoch 52/74\n", "----------\n", "[53] loss: 0.003\n", "Epoch 53/74\n", "----------\n", "[54] loss: 0.001\n", "Epoch 54/74\n", "----------\n", "[55] loss: 0.000\n", "Epoch 55/74\n", "----------\n", "[56] loss: 0.003\n", "Epoch 56/74\n", "----------\n", "[57] loss: 0.005\n", "Epoch 57/74\n", "----------\n", "[58] loss: 0.000\n", "Epoch 58/74\n", "----------\n", "[59] loss: 0.034\n", "Epoch 59/74\n", "----------\n", "[60] loss: 0.039\n", "Epoch 60/74\n", "----------\n", "[61] loss: 0.009\n", "Epoch 61/74\n", "----------\n", "[62] loss: 0.009\n", "Epoch 62/74\n", "----------\n", "[63] loss: 0.000\n", "Epoch 63/74\n", "----------\n", "[64] loss: 0.000\n", "Epoch 64/74\n", "----------\n", "[65] loss: 0.001\n", "Epoch 65/74\n", "----------\n", "[66] loss: 0.019\n", "Epoch 66/74\n", "----------\n", "[67] loss: 0.005\n", "Epoch 67/74\n", "----------\n", "[68] loss: 0.001\n", "Epoch 68/74\n", "----------\n", "[69] loss: 0.000\n", "Epoch 69/74\n", "----------\n", "[70] loss: 0.000\n", "Epoch 70/74\n", "----------\n", "[71] loss: 0.000\n", "Epoch 71/74\n", "----------\n", "[72] loss: 0.000\n", "Epoch 72/74\n", "----------\n", "[73] loss: 0.009\n", "Epoch 73/74\n", "----------\n", "[74] loss: 0.011\n", "Epoch 74/74\n", "----------\n", "[75] loss: 0.001\n", "Finished Training\n" ] } ], "source": [ "import torch.nn as nn\n", "import torch.nn.functional as F\n", "\n", "for epoch in range(num_epochs):\n", " print('Epoch {}/{}'.format(epoch, num_epochs - 1))\n", " print('-' * 10)\n", " \n", " running_loss = 0.0\n", " \n", " for i, (inputs, labels) in enumerate(dataloaders['training']):\n", "# inputs = inputs.to(device)\n", "# labels = labels.to(device)\n", "\n", " # zero the parameter gradients\n", " optimizer.zero_grad()\n", "\n", " # forward + backward + optimize\n", " outputs = net(inputs) # forward\n", " \n", " loss = criterion(outputs, labels) # loss function defined above\n", " loss.backward() # backprop\n", " optimizer.step() \n", "\n", " # print statistics\n", " running_loss += loss.item()\n", " if i % 18 == 0: # print every 18th\n", " print('[%d] loss: %.3f' %\n", " (epoch + 1, running_loss / 18))\n", " running_loss = 0.0\n", "\n", "print('Finished Training')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 5. Test the network on the test data\n", "\n", "We have trained the network for 50 passes over the training dataset. But we need to check if the network has learned anything at all.\n", "\n", "We will check this by predicting the class label that the neural network outputs, and checking it against the ground-truth. If the prediction is correct, we add the sample to the list of correct predictions." ] }, { "cell_type": "code", "execution_count": 7, "metadata": { "scrolled": true }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ "GroundTruth: sadness happiness happiness sadness\n", "Predicted: sadness sadness happiness sadness\n" ] } ], "source": [ "images, labels = next(iter(dataloaders['testing']))\n", "outputs = net(images)\n", "_, predicted = torch.max(outputs, 1)\n", "\n", "# print images\n", "imshow(torchvision.utils.make_grid(images))\n", "\n", "print('GroundTruth: ', ' '.join('%5s' % class_names[labels[j]] for j in range(4)))\n", "\n", "print('Predicted: ', ' '.join('%5s' % class_names[predicted[j]]\n", " for j in range(4)))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### Magic! Way better than I expected! \n", "Let us look at how the network performs on the whole dataset." ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Accuracy of the network on test images: 91 %\n" ] } ], "source": [ "correct = 0\n", "total = 0\n", "with torch.no_grad():\n", " for data in dataloaders['testing']:\n", " images, labels = data\n", " outputs = net(images)\n", " _, predicted = torch.max(outputs.data, 1)\n", " total += labels.size(0)\n", " correct += (predicted == labels).sum().item()\n", "\n", "print('Accuracy of the network on test images: %d %%' % (\n", " 100 * correct / total))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "And, there you have it! In this simple example, we can classify happy and sad faces with super high accuracy! *(better than some people)*" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "Python 3", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.5.2" } }, "nbformat": 4, "nbformat_minor": 2 }