Fracture-Xception.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "Lh-0YEY-zsqO"
      },
      "outputs": [],
      "source": [
        "!mkdir -p ~/.kaggle\n",
        "!cp kaggle.json ~/.kaggle/"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "id": "jwZ32SflMoll"
      },
      "outputs": [],
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "from keras import Sequential\n",
        "from keras import layers\n",
        "from keras.layers import Dense,Conv2D,MaxPooling2D,Flatten,BatchNormalization,Dropout,AveragePooling2D\n",
        "from keras.layers import Activation, Dropout, Flatten, Dense, GaussianNoise, GlobalMaxPooling2D\n",
        "import tensorflow as tf\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "import cv2"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "n9ZDW4-Mz7Of",
        "outputId": "d05b4523-8117-44b8-db3b-0e393481bfef"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Warning: Your Kaggle API key is readable by other users on this system! To fix this, you can run 'chmod 600 /root/.kaggle/kaggle.json'\n",
            "Downloading bone-fracture-detection-using-xrays.zip to /content\n",
            " 99% 170M/172M [00:03<00:00, 43.6MB/s]\n",
            "100% 172M/172M [00:03<00:00, 55.3MB/s]\n"
          ]
        }
      ],
      "source": [
        "#Downloading dataset\n",
        "!kaggle datasets download -d vuppalaadithyasairam/bone-fracture-detection-using-xrays"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "id": "PaxO7oAXz-a0"
      },
      "outputs": [],
      "source": [
        "import zipfile\n",
        "zip_ref = zipfile.ZipFile('/content/bone-fracture-detection-using-xrays.zip', 'r')\n",
        "zip_ref.extractall('/content')\n",
        "zip_ref.close()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_eyE3rRc0EMF",
        "outputId": "479d6526-6b3e-4508-c47b-38ae527dcebb"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Found 8863 files belonging to 2 classes.\n",
            "Found 600 files belonging to 2 classes.\n"
          ]
        }
      ],
      "source": [
        "#Loading the images\n",
        "x_train= keras.utils.image_dataset_from_directory(\n",
        "    directory = '/content/archive (6)/train',\n",
        "    labels='inferred',\n",
        "    label_mode = 'int',\n",
        "    batch_size=100,\n",
        "    image_size=(256,256)\n",
        ")\n",
        "\n",
        "x_test= keras.utils.image_dataset_from_directory(\n",
        "    directory = '/content/archive (6)/val',\n",
        "    labels='inferred',\n",
        "    label_mode = 'int',\n",
        "    batch_size=100,\n",
        "    image_size=(256,256)\n",
        ")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wQQii809Omh_",
        "outputId": "9245b2f7-3e50-4bb2-cbf5-2a48ad9ba0d5"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['fractured', 'not fractured']"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ],
      "source": [
        "class_names = x_train.class_names\n",
        "class_names"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "id": "gxMa58Y-0PwT"
      },
      "outputs": [],
      "source": [
        "# Normalizing\n",
        "def process(image,label):\n",
        "    image = tf.cast(image/255. ,tf.float32)\n",
        "    return image,label\n",
        "\n",
        "x_train = x_train.map(process)\n",
        "x_test = x_test.map(process)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "id": "EYo4lIl2M6aO"
      },
      "outputs": [],
      "source": [
        "input = (256, 256, 3)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "id": "Jnm2v6K28QS2"
      },
      "outputs": [],
      "source": [
        "from tensorflow.keras.applications import Xception\n",
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Dense, GlobalAveragePooling2D"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_VC8ejjSDS7_",
        "outputId": "951dc7ec-d5e5-4bd1-82dd-06f18d24ea90"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/xception/xception_weights_tf_dim_ordering_tf_kernels_notop.h5\n",
            "83683744/83683744 [==============================] - 0s 0us/step\n"
          ]
        }
      ],
      "source": [
        "# Creating the base model\n",
        "base_model = Xception(weights='imagenet', include_top=False)\n",
        "for layer in base_model.layers:\n",
        "    layer.trainable = False"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "id": "RXBEl8q28X-9"
      },
      "outputs": [],
      "source": [
        "model = Sequential()\n",
        "model.add(base_model)\n",
        "model.add(GlobalAveragePooling2D())\n",
        "model.add(Dense(256, activation='relu'))\n",
        "model.add(Dense(1, activation='sigmoid'))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ElJNhexgB1v5",
        "outputId": "678c466b-e4fe-4a19-90dc-42e69bfb6eac"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " xception (Functional)       (None, None, None, 2048   20861480  \n",
            "                             )                                   \n",
            "                                                                 \n",
            " global_average_pooling2d (  (None, 2048)              0         \n",
            " GlobalAveragePooling2D)                                         \n",
            "                                                                 \n",
            " dense (Dense)               (None, 256)               524544    \n",
            "                                                                 \n",
            " dense_1 (Dense)             (None, 1)                 257       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 21386281 (81.58 MB)\n",
            "Trainable params: 524801 (2.00 MB)\n",
            "Non-trainable params: 20861480 (79.58 MB)\n",
            "_________________________________________________________________\n"
          ]
        }
      ],
      "source": [
        "model.summary()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {
        "id": "r2RfYW9a0ih2"
      },
      "outputs": [],
      "source": [
        "model.compile(optimizer=\"adam\", loss=\"binary_crossentropy\", metrics=[\"accuracy\"])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "HHBFNFHCtzLu",
        "outputId": "d25bda28-9e57-4ab1-f348-1883d38a23ea"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/10\n",
            "89/89 [==============================] - 73s 622ms/step - loss: 0.4972 - accuracy: 0.7543 - val_loss: 0.5390 - val_accuracy: 0.6933\n",
            "Epoch 2/10\n",
            "89/89 [==============================] - 52s 575ms/step - loss: 0.2876 - accuracy: 0.8925 - val_loss: 0.5470 - val_accuracy: 0.7267\n",
            "Epoch 3/10\n",
            "89/89 [==============================] - 51s 567ms/step - loss: 0.1915 - accuracy: 0.9405 - val_loss: 0.5517 - val_accuracy: 0.7167\n",
            "Epoch 4/10\n",
            "89/89 [==============================] - 53s 583ms/step - loss: 0.1346 - accuracy: 0.9621 - val_loss: 0.7271 - val_accuracy: 0.7083\n",
            "Epoch 5/10\n",
            "89/89 [==============================] - 50s 554ms/step - loss: 0.0987 - accuracy: 0.9728 - val_loss: 0.5997 - val_accuracy: 0.7200\n",
            "Epoch 6/10\n",
            "89/89 [==============================] - 52s 577ms/step - loss: 0.0674 - accuracy: 0.9845 - val_loss: 0.6347 - val_accuracy: 0.7267\n",
            "Epoch 7/10\n",
            "89/89 [==============================] - 52s 572ms/step - loss: 0.0511 - accuracy: 0.9902 - val_loss: 0.6724 - val_accuracy: 0.7267\n",
            "Epoch 8/10\n",
            "89/89 [==============================] - 52s 566ms/step - loss: 0.0449 - accuracy: 0.9906 - val_loss: 0.7850 - val_accuracy: 0.6950\n",
            "Epoch 9/10\n",
            "89/89 [==============================] - 52s 579ms/step - loss: 0.0308 - accuracy: 0.9962 - val_loss: 0.7294 - val_accuracy: 0.7083\n",
            "Epoch 10/10\n",
            "89/89 [==============================] - 52s 567ms/step - loss: 0.0249 - accuracy: 0.9966 - val_loss: 0.7599 - val_accuracy: 0.7133\n"
          ]
        }
      ],
      "source": [
        "history = model.fit(x_train,epochs=10, batch_size=100, validation_data=x_test)"
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "provenance": [],
      "gpuType": "T4"
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}