Image_Recognition.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Notebook6_starter.ipynb",
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/cs167/notebook-6-niclacey/blob/main/Notebook6_Nicole%20Lacey.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XHNyMbAY0QB5"
      },
      "source": [
        "# Notebook \\# 6\n",
        "Name:\n",
        "Date:\n",
        "\n",
        "## Problem Description:\n",
        "\n",
        "\n",
        "\n",
        "### Hints/ Tips Tricks\n",
        "- Don't forget to enable GPUs under 'Edit', 'Notebook Settings', 'Accerlate Hardware', 'GPU'.\n",
        "- Make sure you unzip the data and put it in your Drive.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "c8YdA-9NIguO",
        "outputId": "a425d391-aa46-4ec5-b9f9-e1e3a3a65e65",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9TyLjTVU0MQr",
        "outputId": "926fe7ef-fc0d-4de9-8379-e532f01e6b6c",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "import keras\n",
        "import tensorflow \n",
        "import sys\n",
        "from matplotlib import pyplot\n",
        "from tensorflow.keras.utils import to_categorical\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Conv2D\n",
        "from keras.layers import MaxPooling2D\n",
        "from keras.layers import Dense\n",
        "from keras.layers import Flatten\n",
        "from tensorflow.keras.optimizers import SGD\n",
        "from keras.preprocessing.image import ImageDataGenerator\n",
        "import sys\n",
        "\n",
        "## Connect to Drive\n",
        "from google.colab import drive\n",
        "import pandas\n",
        "drive.mount('/content/drive')\n",
        "\n",
        "## Image dimensions\n",
        "img_width = 200\n",
        "img_height = 200\n",
        "\n",
        "## Directories for testing and training data\n",
        "train_data_dir = '/content/drive/MyDrive/Datasets/chest_xray/train' \n",
        "test_data_dir = '/content/drive/MyDrive/Datasets/chest_xray/test'\n",
        "\n",
        "## To feed the training images to the neural network in batches of 32 images at a time\n",
        "batch_size = 32\n",
        "\n",
        "## Rescale pixel values from [0, 255] to between 0 and 1\n",
        "datagen = ImageDataGenerator(rescale=1.0/255.0)\n",
        "\n",
        "## Look for training and testing data and figure out the class of each example based on subfolder\n",
        "train_data = datagen.flow_from_directory(\n",
        "        train_data_dir,\n",
        "        target_size=(img_width, img_height),\n",
        "        batch_size=batch_size,\n",
        "        class_mode='binary')\n",
        "\n",
        "test_data = datagen.flow_from_directory(\n",
        "        test_data_dir,\n",
        "        target_size=(img_width, img_height),\n",
        "        batch_size=batch_size,\n",
        "        class_mode='binary')"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n",
            "Found 5216 images belonging to 2 classes.\n",
            "Found 624 images belonging to 2 classes.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "t_i-Cfa41ZvU"
      },
      "source": [
        "# Build your CNN here\n",
        "import warnings\n",
        "warnings.filterwarnings('ignore')\n",
        "\n",
        "model2 = Sequential()\n",
        "model2.add(Conv2D(32, (3, 3), activation = 'relu', input_shape=(img_width, img_height, 3)))\n",
        "model2.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "\n",
        "model2.add(Conv2D(32, (3, 3),activation = 'relu'))\n",
        "model2.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "\n",
        "model2.add(Flatten())\n",
        "model2.add(Dense(64,activation = 'relu'))\n",
        "model2.add(Dense(1,activation='sigmoid'))\n",
        "\n",
        "# need to compile the model before you can use it\n",
        "opt = SGD(lr=0.001, momentum=0.9)\n",
        "model2.compile(optimizer=opt, loss='binary_crossentropy', metrics=['accuracy'])"
      ],
      "execution_count": 19,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XxV4ADsoTmYT",
        "outputId": "5c12a0b5-6855-4386-c708-1bb55c18ff24",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "training_results1 = model2.fit_generator(\n",
        "        train_data, #training set\n",
        "        steps_per_epoch = len(train_data), \n",
        "        epochs=3, #number of epochs \n",
        "        validation_data = test_data, #testing set\n",
        "        validation_steps = len(test_data)\n",
        "        )"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/3\n",
            "163/163 [==============================] - 202s 1s/step - loss: 0.4044 - accuracy: 0.8250 - val_loss: 0.6352 - val_accuracy: 0.7260\n",
            "Epoch 2/3\n",
            "163/163 [==============================] - 201s 1s/step - loss: 0.1720 - accuracy: 0.9339 - val_loss: 0.9611 - val_accuracy: 0.7051\n",
            "Epoch 3/3\n",
            "163/163 [==============================] - 201s 1s/step - loss: 0.2142 - accuracy: 0.9156 - val_loss: 0.3820 - val_accuracy: 0.8189\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QsuuIhVeBVIn"
      },
      "source": [
        "from keras.layers import Dropout\n",
        "from keras.layers import BatchNormalization\n",
        "\n",
        "model3 = Sequential()\n",
        "model3.add(Conv2D(32, (3, 3), activation = 'relu', input_shape=(img_width, img_height, 3)))\n",
        "model3.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "\n",
        "model3.add(Conv2D(64, (3, 3),activation = 'relu'))\n",
        "model3.add(BatchNormalization())\n",
        "model3.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "\n",
        "model3.add(Conv2D(64, (3,3), activation = 'relu'))\n",
        "model3.add(BatchNormalization())\n",
        "model3.add(MaxPooling2D(pool_size= (2,2)))\n",
        "\n",
        "model3.add(Flatten())\n",
        "model3.add(Dense(64,activation = 'relu'))\n",
        "model3.add(Dense(1,activation='sigmoid'))\n",
        "\n",
        "# need to compile the model before you can use it\n",
        "opt = SGD(lr=0.001, momentum=0.9)\n",
        "model3.compile(optimizer=opt, loss='binary_crossentropy', metrics=['accuracy'])\n",
        "early_stopping_cb = tensorflow.keras.callbacks.EarlyStopping(patience=10, restore_best_weights=True)"
      ],
      "execution_count": 13,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eKnjkdQdL-TM",
        "outputId": "f3209b33-5bfc-4589-ea3b-26d6f59e1c61",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "training_results = model3.fit_generator(\n",
        "        train_data, #training set\n",
        "        steps_per_epoch = len(train_data), \n",
        "        epochs=15, #number of epochs \n",
        "        validation_data = test_data, #testing set\n",
        "        validation_steps = len(test_data)\n",
        "        )"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/15\n",
            "163/163 [==============================] - 334s 2s/step - loss: 0.0491 - accuracy: 0.9803 - val_loss: 1.2894 - val_accuracy: 0.6635\n",
            "Epoch 2/15\n",
            "163/163 [==============================] - 334s 2s/step - loss: 0.0278 - accuracy: 0.9896 - val_loss: 1.8804 - val_accuracy: 0.6907\n",
            "Epoch 3/15\n",
            "163/163 [==============================] - 333s 2s/step - loss: 0.0166 - accuracy: 0.9950 - val_loss: 1.1514 - val_accuracy: 0.7500\n",
            "Epoch 4/15\n",
            "163/163 [==============================] - 333s 2s/step - loss: 0.0096 - accuracy: 0.9973 - val_loss: 0.9575 - val_accuracy: 0.7821\n",
            "Epoch 5/15\n",
            "163/163 [==============================] - 333s 2s/step - loss: 0.0066 - accuracy: 0.9988 - val_loss: 0.9418 - val_accuracy: 0.8173\n",
            "Epoch 6/15\n",
            "163/163 [==============================] - 333s 2s/step - loss: 0.0032 - accuracy: 1.0000 - val_loss: 1.3980 - val_accuracy: 0.7853\n",
            "Epoch 7/15\n",
            "163/163 [==============================] - 333s 2s/step - loss: 0.0018 - accuracy: 1.0000 - val_loss: 2.3476 - val_accuracy: 0.7035\n",
            "Epoch 8/15\n",
            "163/163 [==============================] - 333s 2s/step - loss: 0.0014 - accuracy: 1.0000 - val_loss: 1.7253 - val_accuracy: 0.7644\n",
            "Epoch 9/15\n",
            "163/163 [==============================] - 333s 2s/step - loss: 0.0011 - accuracy: 1.0000 - val_loss: 1.6006 - val_accuracy: 0.7772\n",
            "Epoch 10/15\n",
            "163/163 [==============================] - 332s 2s/step - loss: 9.3671e-04 - accuracy: 1.0000 - val_loss: 1.6572 - val_accuracy: 0.7772\n",
            "Epoch 11/15\n",
            "163/163 [==============================] - 333s 2s/step - loss: 8.0475e-04 - accuracy: 1.0000 - val_loss: 1.8859 - val_accuracy: 0.7612\n",
            "Epoch 12/15\n",
            "163/163 [==============================] - 333s 2s/step - loss: 7.4084e-04 - accuracy: 1.0000 - val_loss: 1.9096 - val_accuracy: 0.7612\n",
            "Epoch 13/15\n",
            "163/163 [==============================] - 333s 2s/step - loss: 6.8783e-04 - accuracy: 1.0000 - val_loss: 1.6953 - val_accuracy: 0.7837\n",
            "Epoch 14/15\n",
            "163/163 [==============================] - 334s 2s/step - loss: 6.7518e-04 - accuracy: 1.0000 - val_loss: 1.8898 - val_accuracy: 0.7644\n",
            "Epoch 15/15\n",
            "163/163 [==============================] - 333s 2s/step - loss: 5.6156e-04 - accuracy: 1.0000 - val_loss: 2.0997 - val_accuracy: 0.7548\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "N9FsRvBNweXB",
        "outputId": "704de76b-1a5a-4ea5-c948-49188721ec9d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        }
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline\n",
        "plt.plot(training_results.history['accuracy'])\n",
        "plt.plot(training_results.history['val_accuracy'])\n",
        "plt.title('model accuracy')\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'test'], loc='upper left')\n",
        "plt.show()\n"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "k-FBjnPMLGPv",
        "outputId": "b6e99bde-fbd9-4981-f5c4-a9787ae8e1ab",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        }
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline\n",
        "plt.plot(training_results1.history['accuracy'])\n",
        "plt.plot(training_results1.history['val_accuracy'])\n",
        "plt.title('model accuracy')\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'test'], loc='upper left')\n",
        "plt.show()"
      ],
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "w4Y08U_kNcz8"
      },
      "source": [
        "In the second model I added batch normalization and an extra layer of conv and pool. I also added early stopping because it marginally improved the accuracys (prevents overfitting). I also did a lot more epochs- 15. By the graph I should have stopped at 4 for maximum accuracy so this model does suffer from over fitting. The first model I did doesn't have a peak, which means I could probably have used that one and used a higher epoch. I did experiment with dropout layers but they made the model worse so I ended up leaving it out. They both ended up with very similar highest accuracys (81%). "
      ]
    }
  ]
}