Emotion Recognition.ipynb

emotion-recognition.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Emotion Recognition.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "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": "3.8.3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/gist/Strady123/949f3d3ac8257c1e8e168a08460d1b54/emotion-recognition.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0cxljsVPVJ7S"
      },
      "source": [
        "import numpy as np\n",
        "import torch\n",
        "from torch import nn\n",
        "from torch import optim\n",
        "import torch.nn.functional as F\n",
        "from torchvision import datasets, transforms, models"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "resources": {
            "http://localhost:8080/nbextensions/google.colab/files.js": {
              "data": "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",
              "ok": true,
              "headers": [
                [
                  "content-type",
                  "application/javascript"
                ]
              ],
              "status": 200,
              "status_text": ""
            }
          },
          "base_uri": "https://localhost:8080/",
          "height": 89
        },
        "id": "2xb2XzyAVJ7c",
        "outputId": "83b48d48-9a2e-4ed0-aec2-ae47b2b946c4"
      },
      "source": [
        "from google.colab import files\r\n",
        "\r\n",
        "uploaded = files.upload()\r\n",
        "\r\n",
        "for fn in uploaded.keys():\r\n",
        "      print('User uploaded file \"{name}\" with length {length} bytes'.format(name=fn, length=len(uploaded[fn])))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "\n",
              "     <input type=\"file\" id=\"files-5cffcc47-a517-4d4d-b07e-24dca62500e1\" name=\"files[]\" multiple disabled\n",
              "        style=\"border:none\" />\n",
              "     <output id=\"result-5cffcc47-a517-4d4d-b07e-24dca62500e1\">\n",
              "      Upload widget is only available when the cell has been executed in the\n",
              "      current browser session. Please rerun this cell to enable.\n",
              "      </output>\n",
              "      <script src=\"/nbextensions/google.colab/files.js\"></script> "
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Saving Emotions.zip to Emotions (1).zip\n",
            "User uploaded file \"Emotions.zip\" with length 1449779 bytes\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6_EC8W23oFgV",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "69f9f080-c931-4c09-bbe6-7245042c2448"
      },
      "source": [
        "from zipfile import ZipFile\n",
        "file_name = \"Emotions.zip\"\n",
        "\n",
        "with ZipFile(file_name, 'r') as zip:\n",
        "  zip.extractall()\n",
        "  print(\"Done\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Done\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "r9SPc-kQVJ7h"
      },
      "source": [
        "test_dir = '/content/Emotions/CK+48/test'\n",
        "train_dir = '/content/Emotions/CK+48/train'\n",
        "\n",
        "train_transforms = transforms.Compose([transforms.RandomRotation(30),\n",
        "                                       transforms.RandomResizedCrop(100),\n",
        "                                       transforms.RandomHorizontalFlip(),\n",
        "                                       transforms.ToTensor(),\n",
        "                                       transforms.Normalize([0.5,0.5,0.5],\n",
        "                                                           [0.5,0.5,0.5])])\n",
        "test_transforms = transforms.Compose([transforms.Resize(255),\n",
        "                                      transforms.CenterCrop(224),\n",
        "                                      transforms.ToTensor()])\n",
        "train_data = datasets.ImageFolder(train_dir, transform=train_transforms)\n",
        "trainloader = torch.utils.data.DataLoader(train_data, batch_size=32, shuffle=True)\n",
        "\n",
        "test_data = datasets.ImageFolder(test_dir, transform=test_transforms)\n",
        "testloader = torch.utils.data.DataLoader(test_data, batch_size=32, shuffle=True)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "V3G-tEcGVJ7j",
        "outputId": "73fb8343-f625-4fa9-bc60-675417195038"
      },
      "source": [
        "model = models.alexnet(pretrained=True)\n",
        "model\n",
        "for param in model.parameters():\n",
        "    param.requires_grad = False\n",
        "    \n",
        "classifier = nn.Sequential(nn.Linear(9216, 1000),\n",
        "                                 nn.ReLU(),\n",
        "                                 nn.Dropout(0.2),\n",
        "                                 nn.Linear(1000, 5),\n",
        "                                 nn.LogSoftmax(dim=1))\n",
        "\n",
        "model.classifier = classifier\n",
        "\n",
        "criterion = nn.NLLLoss()\n",
        "\n",
        "optimizer = optim.SGD(model.classifier.parameters(), lr=0.003)\n",
        "model"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "AlexNet(\n",
              "  (features): Sequential(\n",
              "    (0): Conv2d(3, 64, kernel_size=(11, 11), stride=(4, 4), padding=(2, 2))\n",
              "    (1): ReLU(inplace=True)\n",
              "    (2): MaxPool2d(kernel_size=3, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
              "    (3): Conv2d(64, 192, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2))\n",
              "    (4): ReLU(inplace=True)\n",
              "    (5): MaxPool2d(kernel_size=3, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
              "    (6): Conv2d(192, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
              "    (7): ReLU(inplace=True)\n",
              "    (8): Conv2d(384, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
              "    (9): ReLU(inplace=True)\n",
              "    (10): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
              "    (11): ReLU(inplace=True)\n",
              "    (12): MaxPool2d(kernel_size=3, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
              "  )\n",
              "  (avgpool): AdaptiveAvgPool2d(output_size=(6, 6))\n",
              "  (classifier): Sequential(\n",
              "    (0): Linear(in_features=9216, out_features=1000, bias=True)\n",
              "    (1): ReLU()\n",
              "    (2): Dropout(p=0.2, inplace=False)\n",
              "    (3): Linear(in_features=1000, out_features=5, bias=True)\n",
              "    (4): LogSoftmax(dim=1)\n",
              "  )\n",
              ")"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZfvJnxANVJ7m",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a9a6fd29-6ea5-4d47-ba5b-432fa4a717e9"
      },
      "source": [
        "epochs = 8\n",
        "steps = 0\n",
        "running_loss = 0\n",
        "print_every = 5\n",
        "class_number = 5\n",
        "for epoch in range(epochs):\n",
        "    for inputs, labels in trainloader:\n",
        "        steps += 1\n",
        "\n",
        "        \n",
        "        optimizer.zero_grad()\n",
        "        \n",
        "        logps = model.forward(inputs)\n",
        "        loss = criterion(logps, labels)\n",
        "        loss.backward()\n",
        "        optimizer.step()\n",
        "\n",
        "        running_loss += loss.item()\n",
        "        \n",
        "        if steps % print_every == 0:\n",
        "            test_loss = 0\n",
        "            accuracy = 0\n",
        "            model.eval()\n",
        "            with torch.no_grad():\n",
        "                for inputs, labels in testloader:\n",
        "                    \n",
        "                    logps = model.forward(inputs)\n",
        "                    batch_loss = criterion(logps, labels)\n",
        "                    \n",
        "                    test_loss += batch_loss.item()\n",
        "                    \n",
        "                    # Calculate accuracy\n",
        "                    ps = torch.exp(logps)\n",
        "                    top_p, top_class = ps.topk(1, dim=1)\n",
        "                    equals = top_class == labels.view(*top_class.shape)\n",
        "                    accuracy += torch.mean(equals.type(torch.FloatTensor)).item()\n",
        "                    \n",
        "            print(f\"Epoch {epoch+1}/{epochs}.. \"\n",
        "                  f\"Train loss: {running_loss/print_every:.3f}.. \"\n",
        "                  f\"Test loss: {test_loss/len(testloader):.3f}.. \"\n",
        "                  f\"Test accuracy: {accuracy/len(testloader):.3f}\")\n",
        "            running_loss = 0\n",
        "            model.train()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/8.. Train loss: 1.543.. Test loss: 1.578.. Test accuracy: 0.244\n",
            "Epoch 1/8.. Train loss: 1.457.. Test loss: 1.573.. Test accuracy: 0.244\n",
            "Epoch 1/8.. Train loss: 1.422.. Test loss: 1.523.. Test accuracy: 0.244\n",
            "Epoch 2/8.. Train loss: 1.334.. Test loss: 1.515.. Test accuracy: 0.300\n",
            "Epoch 2/8.. Train loss: 1.287.. Test loss: 1.541.. Test accuracy: 0.362\n",
            "Epoch 2/8.. Train loss: 1.242.. Test loss: 1.470.. Test accuracy: 0.350\n",
            "Epoch 2/8.. Train loss: 1.350.. Test loss: 1.475.. Test accuracy: 0.281\n",
            "Epoch 3/8.. Train loss: 1.182.. Test loss: 1.436.. Test accuracy: 0.494\n",
            "Epoch 3/8.. Train loss: 1.213.. Test loss: 1.458.. Test accuracy: 0.500\n",
            "Epoch 3/8.. Train loss: 1.211.. Test loss: 1.507.. Test accuracy: 0.287\n",
            "Epoch 3/8.. Train loss: 1.247.. Test loss: 1.484.. Test accuracy: 0.269\n",
            "Epoch 4/8.. Train loss: 1.208.. Test loss: 1.409.. Test accuracy: 0.588\n",
            "Epoch 4/8.. Train loss: 1.140.. Test loss: 1.401.. Test accuracy: 0.619\n",
            "Epoch 4/8.. Train loss: 1.242.. Test loss: 1.421.. Test accuracy: 0.481\n",
            "Epoch 4/8.. Train loss: 1.122.. Test loss: 1.404.. Test accuracy: 0.656\n",
            "Epoch 5/8.. Train loss: 1.099.. Test loss: 1.431.. Test accuracy: 0.394\n",
            "Epoch 5/8.. Train loss: 1.136.. Test loss: 1.377.. Test accuracy: 0.575\n",
            "Epoch 5/8.. Train loss: 1.066.. Test loss: 1.399.. Test accuracy: 0.569\n",
            "Epoch 5/8.. Train loss: 1.186.. Test loss: 1.351.. Test accuracy: 0.525\n",
            "Epoch 6/8.. Train loss: 1.132.. Test loss: 1.362.. Test accuracy: 0.631\n",
            "Epoch 6/8.. Train loss: 1.064.. Test loss: 1.361.. Test accuracy: 0.481\n",
            "Epoch 6/8.. Train loss: 1.169.. Test loss: 1.359.. Test accuracy: 0.656\n",
            "Epoch 7/8.. Train loss: 0.951.. Test loss: 1.360.. Test accuracy: 0.463\n",
            "Epoch 7/8.. Train loss: 1.101.. Test loss: 1.367.. Test accuracy: 0.600\n",
            "Epoch 7/8.. Train loss: 1.064.. Test loss: 1.352.. Test accuracy: 0.550\n",
            "Epoch 7/8.. Train loss: 1.061.. Test loss: 1.353.. Test accuracy: 0.650\n",
            "Epoch 8/8.. Train loss: 1.064.. Test loss: 1.360.. Test accuracy: 0.606\n",
            "Epoch 8/8.. Train loss: 1.153.. Test loss: 1.343.. Test accuracy: 0.613\n",
            "Epoch 8/8.. Train loss: 0.973.. Test loss: 1.355.. Test accuracy: 0.606\n",
            "Epoch 8/8.. Train loss: 1.012.. Test loss: 1.311.. Test accuracy: 0.606\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZGEqzQSpVJ7n"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5Vc54xUrVJ7o"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}