endoscopy.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": true,
        "id": "psBWCLABkiRP"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "import pickle\n",
        "import cv2\n",
        "import tensorflow\n",
        "from os import listdir\n",
        "from sklearn.preprocessing import LabelBinarizer\n",
        "from tensorflow.keras.models import Sequential, load_model\n",
        "from tensorflow.keras.layers import Conv2D, MaxPooling2D\n",
        "from tensorflow.keras.layers import BatchNormalization\n",
        "from tensorflow.keras.layers import Activation, Flatten, Dropout, Dense\n",
        "from tensorflow.keras import backend as K\n",
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
        "from tensorflow.keras.optimizers import Adam\n",
        "from tensorflow.keras.preprocessing import image\n",
        "from tensorflow.keras.preprocessing.image import img_to_array\n",
        "from sklearn.preprocessing import MultiLabelBinarizer\n",
        "from sklearn.model_selection import train_test_split\n",
        "import matplotlib.pyplot as plt"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "v9rlxB_Yn3sU",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "01be9c4d-2fa9-4f96-cf95-1501835e9995"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Drive mounting: Mounted at /content/drive\n",
            "Unzipping file \"/content/drive/My Drive/labelled_images.zip\" to \"/content/main_input\"\n"
          ]
        }
      ],
      "source": [
        "from google.colab import drive\n",
        "print(\"Drive mounting: \",end=\"\")\n",
        "drive.mount('/content/drive', force_remount=True)\n",
        "\n",
        "import shutil\n",
        "filename = \"/content/drive/My Drive/labelled_images.zip\"\n",
        "extract_dir = \"/content/main_input\"\n",
        "print(f'Unzipping file \"{filename}\" to \"{extract_dir}\"')\n",
        "shutil.unpack_archive(filename, extract_dir)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": true,
        "id": "3X4IkEAskiRR"
      },
      "outputs": [],
      "source": [
        "# Values for the training the model\n",
        "\n",
        "EPOCHS = 25 # default 25\n",
        "INIT_LR = 1e-3\n",
        "BS = 32\n",
        "default_image_size = tuple((256, 256))\n",
        "image_size = 0\n",
        "directory_root = '/content/main_input/labelled_images'\n",
        "width=256\n",
        "height=256\n",
        "depth=3\n",
        "\n",
        "# function for converting image to matrix of the pixel values\n",
        "def convert_image_to_array(image_dir):\n",
        "    try:\n",
        "        image = cv2.imread(image_dir)\n",
        "        # plt.imshow(image)\n",
        "        if image is not None :\n",
        "            image = cv2.resize(image, default_image_size)\n",
        "            return img_to_array(image)\n",
        "        else :\n",
        "            return np.array([])\n",
        "    except Exception as e:\n",
        "        print(f\"Error : {e}\")\n",
        "        return None"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Total dataset <br>\n",
        "name                | no of images <br>\n",
        "<p>-----------------------------</p>\n",
        "Reduced mucosal view | 2906 <br>\n",
        "Foreign body         | 776 <br>\n",
        "Blood - hematin      | 12 <br>\n",
        "Blood - fresh        | 446 <br>\n",
        "Erosion              | 506 <br>\n",
        "Ulcer                | 854 <br>\n",
        "Ileocecal valve      | 4189 <br>\n",
        "Lymphangiectasia     | 592 <br>\n",
        "Pylorus              | 1529 <br>\n",
        "Erythema             | 159 <br>\n",
        "Angiectasia          | 866 <br>\n",
        "Polyp                | 55 <br>\n",
        "Normal clean mucosa  | 34338 <br>\n",
        "Ampulla of vater     | 10"
      ],
      "metadata": {
        "id": "8vKotgGc3h9E"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": true,
        "id": "CXToQwnQkiRR",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "e0f8ce67-bb62-4792-8a4a-2cca1c156919"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Loading images ...\n",
            "Lower range of images per disease: 1000\n",
            "How many images per disease: 500\n",
            "\n",
            "name     |    no of images\n",
            "Lymphangiectasia 592\n",
            "Erosion 506\n",
            "Blood - fresh 446\n",
            "Polyp 55\n",
            "Blood - hematin 12\n",
            "Erythema 159\n",
            "Normal clean mucosa 34338\n",
            "Foreign body 776\n",
            "Reduced mucosal view 2906\n",
            "Angiectasia 866\n",
            "Ampulla of vater 10\n",
            "Ileocecal valve 4189\n",
            "Pylorus 1529\n",
            "Ulcer 854\n",
            "\n",
            "Total images used 2000 from 4 different diseases - ['Ileocecal valve' 'Normal clean mucosa' 'Pylorus' 'Reduced mucosal view']\n"
          ]
        }
      ],
      "source": [
        "image_list, label_list = [], []\n",
        "try:\n",
        "\n",
        "    print(\"Loading images ...\")\n",
        "    filter_count = input('Lower range of images per disease: ')\n",
        "    images_count = input('How many images per disease: ')\n",
        "    print('\\nname     |    no of images')\n",
        "\n",
        "    root_dir = listdir(directory_root)\n",
        "    for folder in root_dir:\n",
        "      label = listdir(f\"{directory_root}/{folder}\")\n",
        "      image_directory = listdir(f\"{directory_root}/{folder}/{label[0]}\")\n",
        "      print(label[0], len(image_directory))\n",
        "      if len(image_directory)>=int(filter_count):\n",
        "        for image in image_directory[:int(images_count)]:\n",
        "          img_path = f\"{directory_root}/{folder}/{label[0]}/{image}\"\n",
        "          image_list.append(convert_image_to_array(img_path))\n",
        "          label_list.append(label[0])\n",
        "except Exception as e:\n",
        "    print(f\"Error : {e}\")\n",
        "\n",
        "label_binarizer = LabelBinarizer()\n",
        "image_labels = label_binarizer.fit_transform(label_list)\n",
        "pickle.dump(label_binarizer,open('label_transform.pkl', 'wb'))\n",
        "n_classes = len(label_binarizer.classes_)\n",
        "print(f\"\\nTotal images used {len(image_list)} from {n_classes} different diseases - {label_binarizer.classes_}\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": true,
        "id": "IUTc7HGQkiRT",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b1b67a3f-57c6-4077-fa8f-c70b49061cf7"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "2000 images normalized\n",
            "Spllited data to train, test : 1300 700 1300 700\n"
          ]
        }
      ],
      "source": [
        "np_image_list = np.array(image_list, dtype=np.float16) / 225.0\n",
        "print(len(np_image_list),'images normalized')\n",
        "x_train, x_test, y_train, y_test = train_test_split(np_image_list, image_labels, test_size=0.35, random_state = 42)\n",
        "# k-fold cross validation here k=1 I think\n",
        "# x_train means evidance or input and y_train means the labels or output\n",
        "print(\"Spllited data to train, test :\",len(x_train),len(x_test),len(y_train),len(y_test))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": true,
        "id": "HVETTeA1kiRT"
      },
      "outputs": [],
      "source": [
        "aug = ImageDataGenerator(\n",
        "    rotation_range=25, width_shift_range=0.1,\n",
        "    height_shift_range=0.1, shear_range=0.2,\n",
        "    zoom_range=0.2,horizontal_flip=True,\n",
        "    fill_mode=\"nearest\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": true,
        "id": "BoqUYMYukiRU",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "69d35931-783a-42ca-e677-8f3dc3652e8e"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_1\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " conv2d_5 (Conv2D)           (None, 256, 256, 32)      896       \n",
            "                                                                 \n",
            " batch_normalization_6 (Bat  (None, 256, 256, 32)      128       \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " max_pooling2d_5 (MaxPoolin  (None, 128, 128, 32)      0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " dropout_4 (Dropout)         (None, 128, 128, 32)      0         \n",
            "                                                                 \n",
            " conv2d_6 (Conv2D)           (None, 128, 128, 64)      18496     \n",
            "                                                                 \n",
            " batch_normalization_7 (Bat  (None, 128, 128, 64)      256       \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " max_pooling2d_6 (MaxPoolin  (None, 64, 64, 64)        0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " conv2d_7 (Conv2D)           (None, 64, 64, 64)        36928     \n",
            "                                                                 \n",
            " batch_normalization_8 (Bat  (None, 64, 64, 64)        256       \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " max_pooling2d_7 (MaxPoolin  (None, 32, 32, 64)        0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " dropout_5 (Dropout)         (None, 32, 32, 64)        0         \n",
            "                                                                 \n",
            " conv2d_8 (Conv2D)           (None, 32, 32, 128)       73856     \n",
            "                                                                 \n",
            " batch_normalization_9 (Bat  (None, 32, 32, 128)       512       \n",
            " chNormalization)                                                \n",
            "                                                                 \n",
            " max_pooling2d_8 (MaxPoolin  (None, 16, 16, 128)       0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " conv2d_9 (Conv2D)           (None, 16, 16, 128)       147584    \n",
            "                                                                 \n",
            " batch_normalization_10 (Ba  (None, 16, 16, 128)       512       \n",
            " tchNormalization)                                               \n",
            "                                                                 \n",
            " max_pooling2d_9 (MaxPoolin  (None, 8, 8, 128)         0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " dropout_6 (Dropout)         (None, 8, 8, 128)         0         \n",
            "                                                                 \n",
            " flatten_1 (Flatten)         (None, 8192)              0         \n",
            "                                                                 \n",
            " dense_2 (Dense)             (None, 1024)              8389632   \n",
            "                                                                 \n",
            " batch_normalization_11 (Ba  (None, 1024)              4096      \n",
            " tchNormalization)                                               \n",
            "                                                                 \n",
            " dropout_7 (Dropout)         (None, 1024)              0         \n",
            "                                                                 \n",
            " dense_3 (Dense)             (None, 4)                 4100      \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 8677252 (33.10 MB)\n",
            "Trainable params: 8674372 (33.09 MB)\n",
            "Non-trainable params: 2880 (11.25 KB)\n",
            "_________________________________________________________________\n"
          ]
        }
      ],
      "source": [
        "# Support Vector Classifier [frpm sklearn.svm import SVC]\n",
        "# Perceptron [from sklearn.linear_model import Perceptron]\n",
        "# Gaussian [from sklearn.naive_bayes import GaussianNB]\n",
        "# K neighbors [from sklearn.neighbors import KNeighborsClassifier] KNeighborsClassifier(n_neighbors=3)\n",
        "# Default model Seuential\n",
        "model = Sequential()\n",
        "inputShape = (height, width, depth)\n",
        "chanDim = -1\n",
        "if K.image_data_format() == \"channels_first\":\n",
        "    inputShape = (depth, height, width)\n",
        "    chanDim = 1\n",
        "\n",
        "model.add(Conv2D(32, (3, 3), activation=\"relu\", padding=\"same\", input_shape=inputShape))\n",
        "model.add(BatchNormalization(axis=chanDim))\n",
        "model.add(MaxPooling2D((2, 2)))\n",
        "model.add(Dropout(0.25))\n",
        "\n",
        "model.add(Conv2D(64, (3, 3), activation=\"relu\", padding=\"same\"))\n",
        "model.add(BatchNormalization(axis=chanDim))\n",
        "model.add(MaxPooling2D((2, 2)))\n",
        "\n",
        "model.add(Conv2D(64, (3, 3), activation=\"relu\", padding=\"same\"))\n",
        "model.add(BatchNormalization(axis=chanDim))\n",
        "model.add(MaxPooling2D((2, 2)))\n",
        "model.add(Dropout(0.25))\n",
        "\n",
        "model.add(Conv2D(128, (3, 3), activation=\"relu\", padding=\"same\"))\n",
        "model.add(BatchNormalization(axis=chanDim))\n",
        "model.add(MaxPooling2D((2, 2)))\n",
        "\n",
        "model.add(Conv2D(128, (3, 3), activation=\"relu\", padding=\"same\"))\n",
        "model.add(BatchNormalization(axis=chanDim))\n",
        "model.add(MaxPooling2D((2, 2)))\n",
        "model.add(Dropout(0.25))\n",
        "\n",
        "model.add(Flatten())\n",
        "model.add(Dense(1024, activation=\"relu\"))\n",
        "model.add(BatchNormalization())\n",
        "model.add(Dropout(0.5))\n",
        "\n",
        "# Replace 'n_classes' with the number of output classes\n",
        "model.add(Dense(n_classes, activation=\"softmax\"))\n",
        "\n",
        "model.summary()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "collapsed": true,
        "id": "wxDxjxokkiRU",
        "outputId": "ee4445c2-65a8-4f90-cf1d-a1e513ed7a60"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Training neural network...\n",
            "Epoch 1/25\n",
            "40/40 [==============================] - 32s 675ms/step - loss: 0.7873 - accuracy: 0.5410 - val_loss: 0.9904 - val_accuracy: 0.2943\n",
            "Epoch 2/25\n",
            "40/40 [==============================] - 22s 559ms/step - loss: 0.5099 - accuracy: 0.6601 - val_loss: 1.4369 - val_accuracy: 0.2543\n",
            "Epoch 3/25\n",
            "40/40 [==============================] - 21s 525ms/step - loss: 0.3927 - accuracy: 0.7232 - val_loss: 0.8878 - val_accuracy: 0.2557\n",
            "Epoch 4/25\n",
            "40/40 [==============================] - 22s 558ms/step - loss: 0.3620 - accuracy: 0.7177 - val_loss: 0.6758 - val_accuracy: 0.2471\n",
            "Epoch 5/25\n",
            "40/40 [==============================] - 21s 521ms/step - loss: 0.3376 - accuracy: 0.7342 - val_loss: 0.7959 - val_accuracy: 0.3157\n",
            "Epoch 6/25\n",
            "40/40 [==============================] - 22s 553ms/step - loss: 0.3122 - accuracy: 0.7666 - val_loss: 0.8388 - val_accuracy: 0.3157\n",
            "Epoch 7/25\n",
            "40/40 [==============================] - 21s 529ms/step - loss: 0.3227 - accuracy: 0.7508 - val_loss: 1.4102 - val_accuracy: 0.3186\n",
            "Epoch 8/25\n",
            "40/40 [==============================] - 23s 563ms/step - loss: 0.2926 - accuracy: 0.7642 - val_loss: 1.5170 - val_accuracy: 0.3071\n",
            "Epoch 9/25\n",
            "40/40 [==============================] - 22s 540ms/step - loss: 0.2809 - accuracy: 0.7823 - val_loss: 1.4961 - val_accuracy: 0.3371\n",
            "Epoch 10/25\n",
            "40/40 [==============================] - 22s 553ms/step - loss: 0.2649 - accuracy: 0.7973 - val_loss: 1.4321 - val_accuracy: 0.3343\n",
            "Epoch 11/25\n",
            "40/40 [==============================] - 21s 515ms/step - loss: 0.2486 - accuracy: 0.8218 - val_loss: 0.7627 - val_accuracy: 0.4100\n",
            "Epoch 12/25\n",
            "40/40 [==============================] - 22s 552ms/step - loss: 0.2589 - accuracy: 0.8052 - val_loss: 0.4923 - val_accuracy: 0.6514\n",
            "Epoch 13/25\n",
            "40/40 [==============================] - 23s 569ms/step - loss: 0.2587 - accuracy: 0.8178 - val_loss: 0.5393 - val_accuracy: 0.5343\n",
            "Epoch 14/25\n",
            "40/40 [==============================] - 22s 553ms/step - loss: 0.2352 - accuracy: 0.8107 - val_loss: 0.5815 - val_accuracy: 0.5814\n",
            "Epoch 15/25\n",
            "40/40 [==============================] - 21s 519ms/step - loss: 0.2204 - accuracy: 0.8431 - val_loss: 0.4109 - val_accuracy: 0.6643\n",
            "Epoch 16/25\n",
            "40/40 [==============================] - 22s 547ms/step - loss: 0.2183 - accuracy: 0.8352 - val_loss: 0.4466 - val_accuracy: 0.6557\n",
            "Epoch 17/25\n",
            "40/40 [==============================] - 22s 554ms/step - loss: 0.2309 - accuracy: 0.8194 - val_loss: 0.3996 - val_accuracy: 0.7243\n",
            "Epoch 18/25\n",
            "40/40 [==============================] - 22s 557ms/step - loss: 0.2228 - accuracy: 0.8344 - val_loss: 0.4791 - val_accuracy: 0.7157\n",
            "Epoch 19/25\n",
            "40/40 [==============================] - 22s 555ms/step - loss: 0.2062 - accuracy: 0.8509 - val_loss: 0.5425 - val_accuracy: 0.5943\n",
            "Epoch 20/25\n",
            "40/40 [==============================] - 21s 526ms/step - loss: 0.2134 - accuracy: 0.8352 - val_loss: 0.8512 - val_accuracy: 0.5214\n",
            "Epoch 21/25\n",
            "40/40 [==============================] - 22s 555ms/step - loss: 0.1941 - accuracy: 0.8486 - val_loss: 0.3134 - val_accuracy: 0.7743\n",
            "Epoch 22/25\n",
            "40/40 [==============================] - 22s 550ms/step - loss: 0.2002 - accuracy: 0.8502 - val_loss: 0.5497 - val_accuracy: 0.6386\n",
            "Epoch 23/25\n",
            "40/40 [==============================] - 21s 531ms/step - loss: 0.2040 - accuracy: 0.8533 - val_loss: 0.6300 - val_accuracy: 0.6371\n",
            "Epoch 24/25\n",
            "40/40 [==============================] - 22s 550ms/step - loss: 0.1948 - accuracy: 0.8415 - val_loss: 1.5548 - val_accuracy: 0.4286\n",
            "Epoch 25/25\n",
            "40/40 [==============================] - 21s 535ms/step - loss: 0.1760 - accuracy: 0.8778 - val_loss: 0.5040 - val_accuracy: 0.6586\n"
          ]
        }
      ],
      "source": [
        "model.compile(loss=\"binary_crossentropy\", optimizer=\"adam\", metrics=[\"accuracy\"])\n",
        "print(\"Training neural network...\")\n",
        "history = model.fit(\n",
        "    aug.flow(x_train, y_train, batch_size=BS),\n",
        "    validation_data=(x_test, y_test),\n",
        "    steps_per_epoch=len(x_train) // BS,\n",
        "    epochs=EPOCHS, verbose=1\n",
        "    )"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "AdToMc2j8SE9",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "4278a645-1cad-4204-d0e4-5d24fee5b65f"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "{'verbose': 1, 'epochs': 25, 'steps': 40}\n",
            "dict_keys(['loss', 'accuracy', 'val_loss', 'val_accuracy'])\n"
          ]
        }
      ],
      "source": [
        "# [ For checking only ]\n",
        "print(history.params)\n",
        "print(history.history.keys())"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "2MfOB09akiRU",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 887
        },
        "outputId": "d91d756f-4245-4ac4-9162-a7faa0bba49f"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "acc = history.history['accuracy']\n",
        "val_acc = history.history['val_accuracy']\n",
        "loss = history.history['loss']\n",
        "val_loss = history.history['val_loss']\n",
        "epochs = range(1, len(acc) + 1)\n",
        "#Train and validation accuracy\n",
        "plt.plot(epochs, acc, 'b', label='Training accurarcy')\n",
        "plt.plot(epochs, val_acc, 'r', label='Validation accurarcy')\n",
        "plt.title('Training and Validation accurarcy')\n",
        "plt.legend()\n",
        "\n",
        "plt.figure()\n",
        "#Train and validation loss\n",
        "plt.plot(epochs, loss, 'b', label='Training loss')\n",
        "plt.plot(epochs, val_loss, 'r', label='Validation loss')\n",
        "plt.title('Training and Validation loss')\n",
        "plt.legend()\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": true,
        "id": "M8SVhwYUkiRU",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "8a782ce6-ce94-4519-c812-31b3a0544499"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model saved\n",
            "Calculating model accuracy:\n",
            "22/22 [==============================] - 2s 33ms/step - loss: 0.5040 - accuracy: 0.6586\n",
            "Test Accuracy: 66%\n"
          ]
        }
      ],
      "source": [
        "model.save(\"endoscopy.keras\")\n",
        "print('Model saved')\n",
        "print(\"Calculating model accuracy:\")\n",
        "loaded_model = load_model(\"endoscopy.keras\")\n",
        "scores = loaded_model.evaluate(x_test, y_test)\n",
        "print(f\"Test Accuracy: {round(scores[1]*100)}%\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": true,
        "id": "5MaIKDCvkiRU",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "223e6de1-45ec-4268-a0f1-1aaf2e97412a"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 1s 753ms/step\n",
            "[[8.3510022e-06 3.1811255e-04 2.8580211e-03 2.2957191e-05 8.4039085e-02\n",
            "  1.9869252e-05 8.7154651e-05 2.4789264e-02 8.8707161e-01 7.8556308e-04]]\n",
            "probability:0.8870716\n",
            "Reduced mucosal view\n"
          ]
        }
      ],
      "source": [
        "image_location = f\"/content/main_input/labelled_images/reduced_mucosal_view/Reduced mucosal view/8ebf0e483cac48d6_15375.jpg\"\n",
        "im=convert_image_to_array(image_location)\n",
        "np_image_li = np.array(im, dtype=np.float16) / 225.0\n",
        "npp_image = np.expand_dims(np_image_li, axis=0)\n",
        "result = loaded_model.predict(npp_image)\n",
        "print(result)\n",
        "itemindex = np.where(result==np.max(result))\n",
        "print(\"probability:\"+str(np.max(result))+\"\\n\"+label_binarizer.classes_[itemindex[1][0]])"
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "gpuType": "T4",
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "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": "3.10.12"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}