capture_img.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "capture_img.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": "3.8.8"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/MizanMustakim/Deepfake_Detection/blob/main/capture_img.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "m_vWZUJq-BhQ"
      },
      "source": [
        "# **Deepfake Detection Using CNN**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "N0VLHj_d-NhS"
      },
      "source": [
        "**Importing all necessary libraries**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3cIpNN2aG8Q5"
      },
      "source": [
        "import dlib\n",
        "import cv2\n",
        "import os\n",
        "import json\n",
        "from PIL import Image, ImageChops, ImageEnhance\n",
        "\n",
        "import tensorflow as tf\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sn\n",
        "import pandas as pd\n",
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator, img_to_array, load_img\n",
        "from tensorflow.keras.utils import to_categorical\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.metrics import confusion_matrix\n",
        "\n",
        "import keras_preprocessing\n",
        "from keras_preprocessing import image\n",
        "import shutil\n",
        "\n",
        "\n",
        "from tensorflow.keras.models import Model"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hK8zklXD-65_"
      },
      "source": [
        "**Mount the drive to colab**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0y1iClRCJbpJ",
        "outputId": "7da01983-ef10-4468-c0a3-985cd51d365b"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "B8EGiij2_Hpv"
      },
      "source": [
        "**Capturing images from the vidoes. Using the metadata, splitting the images into two labels named \"fake\" and \"real\".**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MXSRededG8Q6"
      },
      "source": [
        "train_frame_folder = '/content/drive/MyDrive/DATA/dfdc_train_part_05/dfdc_train_part_5'\n",
        "\n",
        "meta_path = \"\"\n",
        "\n",
        "for i in os.listdir(train_frame_folder):\n",
        "    if i.split(\".\")[1] == \"json\":\n",
        "        meta_path = train_frame_folder + \"/\" + i\n",
        "\n",
        "with open(os.path.join(train_frame_folder, meta_path), 'r') as file:\n",
        "    data = json.load(file)\n",
        "list_of_train_data = [f for f in os.listdir(train_frame_folder) if f.endswith('.mp4')]\n",
        "detector = dlib.get_frontal_face_detector()\n",
        "for vid in list_of_train_data:\n",
        "    count = 0\n",
        "    cap = cv2.VideoCapture(os.path.join(train_frame_folder, vid))\n",
        "    frameRate = cap.get(5)\n",
        "    while cap.isOpened():\n",
        "        frameId = cap.get(1)\n",
        "        ret, frame = cap.read()\n",
        "        if ret != True:\n",
        "            break\n",
        "        if frameId % ((int(frameRate)+1)*1) == 0:\n",
        "            face_rects, scores, idx = detector.run(frame, 0)\n",
        "            for i, d in enumerate(face_rects):\n",
        "                x1 = d.left()\n",
        "                y1 = d.top()\n",
        "                x2 = d.right()\n",
        "                y2 = d.bottom()\n",
        "                crop_img = frame[y1:y2, x1:x2]\n",
        "                if data[vid]['label'] == 'REAL':\n",
        "                    cv2.imwrite('/content/drive/MyDrive/DATA/DeepVideos/test_data/real/'+vid.split('.')[0]+'_'+str(count)+'.jpg', frame)\n",
        "                elif data[vid]['label'] == 'FAKE':\n",
        "                    cv2.imwrite('/content/drive/MyDrive/DATA/DeepVideos/test_data/fake/'+vid.split('.')[0]+'_'+str(count)+'.jpg', frame) \n",
        "                    # print(count+1)\n",
        "                count+=1"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GqM3AsLW_p-n"
      },
      "source": [
        "**There are showing two ways to preprocess the data before training.** \n",
        "\n",
        "One way is to convert the images to grayscale, and store them into csv file. And the another way is using ImageDataGenerator library."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "e-WGzJ9PBDye"
      },
      "source": [
        "**Gray Scale method**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4Xkob9tQ7qTK"
      },
      "source": [
        "input_shape = (28,28, 3)\n",
        "data_dir = '/content/drive/MyDrive/DATA/DeepVideos'\n",
        "\n",
        "real_data = [f for f in os.listdir(data_dir+'/real') if f.endswith('.jpg')]\n",
        "fake_data = [f for f in os.listdir(data_dir+'/fake') if f.endswith('.jpg')]\n",
        "\n",
        "X = []\n",
        "Y = []\n",
        "\n",
        "for img in real_data:\n",
        "    image = Image.open(data_dir+'/real/'+img)\n",
        "    resized_img = image.resize((28,28))\n",
        "    X.append(img_to_array(resized_img).flatten() / 255.0)\n",
        "    # X.append(img_to_array(image).flatten() / 255.0)\n",
        "    Y.append(1)\n",
        "for img in fake_data:\n",
        "    image = Image.open(data_dir+'/fake/'+img)\n",
        "    resized_img = image.resize((28,28))\n",
        "    X.append(img_to_array(resized_img).flatten() / 255.0)\n",
        "    # X.append(img_to_array(image).flatten() / 255.0)\n",
        "    Y.append(0)\n",
        "\n",
        "Y_val_org = Y\n",
        "\n",
        "#Normalization\n",
        "X = np.array(X)\n",
        "Y = to_categorical(Y, 2)\n",
        "\n",
        "#Reshape\n",
        "X = X.reshape(-1, 28, 28, 3)\n",
        "\n",
        "np.savetxt('/content/drive/MyDrive/DATA/DeepVideos/X.csv', X.reshape(3,-1), delimiter=' ')\n",
        "np.savetxt('/content/drive/MyDrive/DATA/DeepVideos/y.csv', Y, delimiter=' ')\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KhhGVSj6ZoIF"
      },
      "source": [
        "X = np.loadtxt(\"/content/drive/MyDrive/DATA/DeepVideos/X.csv\")\n",
        "X=X.reshape(-1,28,28,3)\n",
        "\n",
        "Y = np.loadtxt(\"/content/drive/MyDrive/DATA/DeepVideos/y.csv\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "J5PeCqZ1DwHW"
      },
      "source": [
        "If we use the above method, then we can split our datasets like this way."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8sEQfjP-CLrV"
      },
      "source": [
        "#Train-Test split\n",
        "X_train, X_val, Y_train, Y_val = train_test_split(X, Y, test_size = 0.3, random_state=5)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "STy0CCCLBXSP"
      },
      "source": [
        "**Using ImageDataGenerator Library**\n",
        "\n",
        "In this method, we just have to specify the directory path of our data. Then the library will process the images by itself. In addtion, to get a better accuracy I used ***Image Augmentation*** here, so that our model can testify our test data with less loss including more accuracy."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Fw11VVTx3ooU",
        "outputId": "7f74a79c-dccc-484d-feae-6ca63ba4aacf"
      },
      "source": [
        "TRAINING_DIR = \"/content/drive/MyDrive/DATA/DeepVideos/test_data\"\n",
        "training_datagen = ImageDataGenerator(\n",
        "      rescale = 1./255,\n",
        "      rotation_range=40,\n",
        "      width_shift_range=0.2,\n",
        "      height_shift_range=0.2,\n",
        "      shear_range=0.2,\n",
        "      zoom_range=0.2,\n",
        "      horizontal_flip=True)\n",
        "\n",
        "VALIDATION_DIR = \"/content/drive/MyDrive/DATA/DeepVideos/train_Data\"\n",
        "validation_datagen = ImageDataGenerator(rescale = 1./255)\n",
        "\n",
        "train_generator = training_datagen.flow_from_directory(\n",
        "\tTRAINING_DIR,\n",
        "\ttarget_size=(48,48),\n",
        "\tclass_mode='binary',\n",
        "  batch_size= 300\n",
        ")\n",
        "\n",
        "validation_generator = validation_datagen.flow_from_directory(\n",
        "\tVALIDATION_DIR,\n",
        "\ttarget_size=(48,48),\n",
        "\tclass_mode='binary',\n",
        "  batch_size= 300\n",
        ")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 17879 images belonging to 2 classes.\n",
            "Found 6333 images belonging to 2 classes.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bM-ktxU8CTMu"
      },
      "source": [
        "This method is used to get the label indices after using **ImageDataGenerator**. Here is shown only for training data, similarly we can get for the validation data."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Q4zA4cS13sZv",
        "outputId": "dbefc2ef-2c27-4e94-fd85-f2ac94e26d78"
      },
      "source": [
        "train_generator.class_indices"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'fake': 0, 'real': 1}"
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "M57dyKLbC9je"
      },
      "source": [
        "**Another way to split datasets.**\n",
        "\n",
        "The next cell is not necessary all the time. If we have less data, then we can split our data into two sets like this way. One is for training, and the another is for validation."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "32RFmqPz5IBi"
      },
      "source": [
        "root_dir = '/content/drive/MyDrive/DATA/DeepVideos/Data/'\n",
        "classes_dir = ['fake', 'real']\n",
        "\n",
        "test_ratio = 0.20\n",
        "\n",
        "for cls in classes_dir:\n",
        "    os.makedirs(root_dir +'train/' + cls)\n",
        "    os.makedirs(root_dir +'test/' + cls)\n",
        "\n",
        "    src = root_dir + cls\n",
        "\n",
        "    allFileNames = os.listdir(src)\n",
        "    np.random.shuffle(allFileNames)\n",
        "    train_FileNames, test_FileNames = np.split(np.array(allFileNames),\n",
        "                                                              [int(len(allFileNames)* (1 - test_ratio))])\n",
        "\n",
        "\n",
        "    train_FileNames = [src+'/'+ name for name in train_FileNames.tolist()]\n",
        "    test_FileNames = [src+'/' + name for name in test_FileNames.tolist()]\n",
        "\n",
        "    print(\"*****************************\")\n",
        "    print('Total images: ', len(allFileNames))\n",
        "    print('Training: ', len(train_FileNames))\n",
        "    print('Testing: ', len(test_FileNames))\n",
        "    print(\"*****************************\")\n",
        "\n",
        "\n",
        "    for name in train_FileNames:\n",
        "        for i in classes_dir:\n",
        "            shutil.copy(name, root_dir +'train/' + i)\n",
        "\n",
        "    for name in test_FileNames:\n",
        "        for i in classes_dir:\n",
        "            shutil.copy(name, root_dir +'test/' + i)\n",
        "    print(\"Copying Done!\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3ymbpRovD9xn"
      },
      "source": [
        "**Building the model.**\n",
        "\n",
        "This model is about the **Convolutional Neural Network** (CNN). Here I used 3 convolutional layers with 3 maxpooling2D layers. Then we flatten the outputs coming from the CNN layers, and feed them into the **Deep Neural Network** (DNN) layers. Our model is working as **Binary Classification** model."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gYR5dmfwonPg",
        "outputId": "c7d14b98-9f12-49fe-8e00-625859b4b62f"
      },
      "source": [
        "model1 = tf.keras.models.Sequential([\n",
        "    # Note the input shape is the desired size of the image 150x150 with 3 bytes color\n",
        "    # This is the first convolution\n",
        "    tf.keras.layers.Conv2D(64, (3,3), activation='relu', input_shape=(48, 48, 3)),\n",
        "    tf.keras.layers.MaxPooling2D(2, 2),\n",
        "    # The second convolution\n",
        "    tf.keras.layers.Conv2D(64, (3,3), activation='relu'),\n",
        "    tf.keras.layers.MaxPooling2D(2,2),\n",
        "    # The third convolution\n",
        "    tf.keras.layers.Conv2D(128, (3,3), activation='relu'),\n",
        "    tf.keras.layers.MaxPooling2D(2,2),\n",
        "    # Flatten the results to feed into a DNN\n",
        "    tf.keras.layers.Flatten(),\n",
        "    # tf.keras.layers.Dropout(0.2),\n",
        "    # 512 neuron hidden layer\n",
        "    tf.keras.layers.Dense(16, activation='relu'),\n",
        "    tf.keras.layers.Dense(1, activation='sigmoid')\n",
        "])\n",
        "\n",
        "\n",
        "model1.summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_1\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_3 (Conv2D)            (None, 46, 46, 64)        1792      \n",
            "_________________________________________________________________\n",
            "max_pooling2d_3 (MaxPooling2 (None, 23, 23, 64)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_4 (Conv2D)            (None, 21, 21, 64)        36928     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_4 (MaxPooling2 (None, 10, 10, 64)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_5 (Conv2D)            (None, 8, 8, 128)         73856     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_5 (MaxPooling2 (None, 4, 4, 128)         0         \n",
            "_________________________________________________________________\n",
            "flatten_1 (Flatten)          (None, 2048)              0         \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 16)                32784     \n",
            "_________________________________________________________________\n",
            "dense_3 (Dense)              (None, 1)                 17        \n",
            "=================================================================\n",
            "Total params: 145,377\n",
            "Trainable params: 145,377\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "I3YBBgCzGIXG"
      },
      "source": [
        "Now, compiled the model."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Vw4_jV3vpkSc"
      },
      "source": [
        "model1.compile(loss='binary_crossentropy',\n",
        "              optimizer = \"adam\",\n",
        "              metrics=['accuracy'])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-6rVXONZGOda"
      },
      "source": [
        "In this cell, the model will be trained, and will be saved to the drive directory where I want to save.."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JsE7Etxs8T-h",
        "outputId": "28cb8e1c-561e-4b06-f57a-5418f6b8a068"
      },
      "source": [
        "#Currently not used\n",
        "# early_stopping = EarlyStopping(monitor='val_loss',\n",
        "#                                min_delta=0,\n",
        "#                                patience=2,\n",
        "#                                verbose=0, mode='auto')\n",
        "EPOCHS = 5\n",
        "# BATCH_SIZE = 300\n",
        "history = model1.fit(train_generator, epochs = EPOCHS, steps_per_epoch=10, validation_data= validation_generator, verbose = 1)\n",
        "\n",
        "model1.save('/content/drive/MyDrive/DATA/DeepVideos/deepfake-detection-model.h5')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/5\n",
            "10/10 [==============================] - 956s 100s/step - loss: 0.4573 - accuracy: 0.8517 - val_loss: 0.3391 - val_accuracy: 0.9195\n",
            "Epoch 2/5\n",
            "10/10 [==============================] - 574s 60s/step - loss: 0.4115 - accuracy: 0.8480 - val_loss: 0.3241 - val_accuracy: 0.9195\n",
            "Epoch 3/5\n",
            "10/10 [==============================] - 542s 57s/step - loss: 0.3817 - accuracy: 0.8647 - val_loss: 0.3641 - val_accuracy: 0.9195\n",
            "Epoch 4/5\n",
            "10/10 [==============================] - 556s 58s/step - loss: 0.3972 - accuracy: 0.8437 - val_loss: 0.3729 - val_accuracy: 0.9195\n",
            "Epoch 5/5\n",
            "10/10 [==============================] - 551s 57s/step - loss: 0.3596 - accuracy: 0.8547 - val_loss: 0.3936 - val_accuracy: 0.9195\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SblmYdeCHo24"
      },
      "source": [
        "Now, we will visualize the accuray and the loss value on the training data and the validation data what we got during the training."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YyfHI8y58cy9",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 300
        },
        "outputId": "58ede740-1d81-4ff8-ab44-13027e4b6516"
      },
      "source": [
        "f, (ax1, ax2) = plt.subplots(1, 2, figsize=(20, 4))\n",
        "t = f.suptitle('CNN training with Fine-Tuning & Image Augmentation Performance ', fontsize=12)\n",
        "f.subplots_adjust(top=0.85, wspace=0.3)\n",
        "\n",
        "epoch_list = list(range(1,EPOCHS+1))\n",
        "ax1.plot(epoch_list, history.history['accuracy'], label='Train Accuracy')\n",
        "ax1.plot(epoch_list, history.history['val_accuracy'], label='Validation Accuracy')\n",
        "ax1.set_xticks(np.arange(0, EPOCHS+1, 1))\n",
        "ax1.set_ylabel('Accuracy Value')\n",
        "ax1.set_xlabel('Epoch #')\n",
        "ax1.set_title('Accuracy')\n",
        "l1 = ax1.legend(loc=\"best\")\n",
        "\n",
        "ax2.plot(epoch_list, history.history['loss'], label='Train Loss')\n",
        "ax2.plot(epoch_list, history.history['val_loss'], label='Validation Loss')\n",
        "ax2.set_xticks(np.arange(0, EPOCHS+1, 1))\n",
        "ax2.set_ylabel('Loss Value')\n",
        "ax2.set_xlabel('Epoch #')\n",
        "ax2.set_title('Loss')\n",
        "l2 = ax2.legend(loc=\"best\")\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1440x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PGRneFzuGxv-"
      },
      "source": [
        "**Loading the saved model**\n",
        "\n",
        "Now, just to testify we will load our previously saved model, and make prediction on the test data. In addition, after getting the predicted values we will produce a **confusion matrix**."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RzKQ0Y8mb2sO"
      },
      "source": [
        "model_load = tf.keras.models.load_model(\"/content/drive/MyDrive/DATA/DeepVideos/deepfake-detection-model.h5\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 403
        },
        "id": "fwUdgQVF3ver",
        "outputId": "048672c4-d866-43be-f71d-21e9db647f2e"
      },
      "source": [
        "#Output confusion matrix\n",
        "def print_confusion_matrix(y_true, y_pred):\n",
        "    cm = confusion_matrix(y_true, y_pred)\n",
        "    print('True positive = ', cm[0][0])\n",
        "    print('False positive = ', cm[0][1])\n",
        "    print('False negative = ', cm[1][0])\n",
        "    print('True negative = ', cm[1][1])\n",
        "    print('\\n')\n",
        "    df_cm = pd.DataFrame(cm, range(2), range(2))\n",
        "    sn.set(font_scale=1.4) # for label size\n",
        "    sn.heatmap(df_cm, annot=True, annot_kws={\"size\": 16}) # font size\n",
        "    plt.ylabel('Actual label', size = 20)\n",
        "    plt.xlabel('Predicted label', size = 20)\n",
        "    plt.xticks(np.arange(2), ['Fake', 'Real'], size = 16)\n",
        "    plt.yticks(np.arange(2), ['Fake', 'Real'], size = 16)\n",
        "    plt.ylim([2, 0])\n",
        "    plt.show()\n",
        "\n",
        "\n",
        "imag_pred = []\n",
        "y = []\n",
        "\n",
        "for fn in os.listdir(\"/content/drive/MyDrive/DATA/DeepVideos/train_Data\"):\n",
        "  # predicting images\n",
        "  for i in os.listdir(\"/content/drive/MyDrive/DATA/DeepVideos/train_Data/\" + fn):\n",
        "    if fn == \"fake\":\n",
        "      y.append(0)\n",
        "    elif fn == \"real\":\n",
        "      y.append(1)\n",
        "    path = \"/content/drive/MyDrive/DATA/DeepVideos/train_Data/\" + fn + \"/\" + i\n",
        "    img = image.load_img(path, target_size=(48,48))\n",
        "    x = image.img_to_array(img)\n",
        "    x = np.expand_dims(x, axis=0)\n",
        "\n",
        "    images = np.vstack([x])\n",
        "    \n",
        "    imag_pred.append(round(model_load.predict(images)[0][0]))\n",
        "\n",
        "    \n",
        "print_confusion_matrix(y, imag_pred)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "True positive =  5823\n",
            "False positive =  0\n",
            "False negative =  510\n",
            "True negative =  0\n",
            "\n",
            "\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZUAAAEdCAYAAADEuPqOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzdeVyU1f7A8c+wugDiioJgIQiCaJCKS7hrblzTyjQXcL3upuaWyy0rLcrQFM2FvK5pmNrV1DQ1cMtuol33ZVBxIVwQBhQYlvn9wY/JkW0GBhnx+76v5/VzznOe83xH+vnlPOc85yg0Go0GIYQQwgjMyjoAIYQQ5YckFSGEEEYjSUUIIYTRSFIRQghhNJJUhBBCGI0kFSGEEEYjSUUIIYTRSFIRQghhNJJUhBBCGI1FWQfwtISEBP78808SExNp37499vb2pKenY2lpiZmZ5EAhhDBlJpNUNBoNISEhbNiwgYyMDBQKBVu3bsXe3p4xY8bg5+fH2LFjyzpMIYQQhTCZX/1XrFjBxo0bGTt2LN9//z1PLknWvn17fv3117ILTgghhF5MpqcSERHB2LFj+ec//0lWVpbOORcXF2JjYw1uM+N+jLHCE+VERceAsg5BmKhM9e0SXW/IvzeWNVxLdC9TZjJJJT4+niZNmuR7ztLSktTU1GcckRBCGCA7q+g6LwCTefzl4ODAlStX8j136dIl6tat+4wjEkIIA2iy9T/KMZNJKl27diUsLIyTJ09qyxQKBdeuXePbb7+le/fuZRidEEIUITtb/6McU5jKJl1paWkMHTqUU6dO4ejoyO3bt3F2diYuLg5fX1/Cw8OxsrIyqE0ZUxFPkzEVUZCSjqmo75zTu66Vo3eJ7mXKTCapAGRlZbFz506OHDlCQkIC9vb2BAQEEBgYiIWF4cM/klTE0ySpiIKUOKnc/FPvulbO+Y8flwcmk1SOHz9Oy5YtCzwfHh7OsGHDDGpTkop4miQVUZASJ5Ub0XrXtarnV6J7mTKTGVMZP348Fy9ezPfcmjVrWLhw4TOOSAghDCAD9YAJJZVu3boxfPhwbt26pVP+73//m5CQEGbNmlVGkQkhhB5koB4woaTy0Ucf0bhxY4YOHUpCQgIA69at47PPPmPmzJkMGDCgjCMUQoiCaTTZeh/lmcmMqQCkp6cTHBxMeno63bt3Z+HChUyfPp3g4OBitSdjKuJpMqYiClLSMZX0K8f0rmvt3qpE9zJlJtNTAbC2tmbFihWo1WoWLlzI1KlTi51QhBDimcrK0P8ox8p0mZZp06blW169enUePHjAxYsXtXUUCgWff/75swxPCCH0V84fa+mrTJPKH3/8UeC5ihUr6pxXKBTPIiQhhCiecj4Ar68yTSoHDx4sy9sLIYTxSE8FMKFVioUQ4rkmPRXARJPKgwcPSE9Pz1Pu6OhYBtEIIUTRNNnlewBeXyaTVLKzs1m0aBFbtmxBpVLlW+fChQvPOCohhNCT9FQAE5pSvHbtWjZu3MiQIUPQaDT885//ZPTo0dStWxcXFxc+/vjjsg5RCCEKJsu0ACaUVLZt28bYsWMZMWIEAJ07d2bChAns3r2bWrVqERcXV8YRCiFEIbKz9D/KMZNJKjdv3qRRo0aYm5tjYWFBWloakLOVcFBQED/88EMZRyiEEIWQngpgQknFxsZGOzhfq1Ytrl27pj2XlZVFUlJSWYUmhBBFkwUlARMaqPfy8kKpVBIQEMBrr73GkiVLqFChAubm5ixatAgvL6+yDlEIIQqWlVnWEZiEMk0qx48fp3HjxlSuXJmgoCBu3rwJ5Oytcu7cOd5//30gZyrxnDlzyjJUIYQoXDnvgeirTJPK0KFD2bJlC40bN6Z169ZkZ2czYMAAPv30U7Zu3UpsbCypqanUr18fS0vLsgxVCCEKpdGU7wF4fZXpmMrTq+5rNBpOnjzJo0ePUCgU1KtXD09PT0koQgjTV0pjKtu2bcPDwyPPMW/ePJ16kZGR9O7dGx8fHzp16sT69evzbS88PJwOHTrQuHFj+vTpw/Hjx/PUSUlJYe7cufj7++Pr68uoUaPybKBYEJMZUxFCiOdaKc/qWr16Nba2ttrPNWrU0P751KlTjBkzhl69ejF9+nSio6OZP38+FhYW9O/fX1svPDyc0NBQJk2ahJeXFxEREYwcOZKIiAg8PT219aZMmcK5c+eYM2cONjY2fP311wQHB7Nz504qVqxYaJySVIQQwhhKeUzF29ubatWq5XsuLCwMLy8v5s+fD0CLFi2Ii4sjLCyMd955BzMzM9RqNcuXL2fw4MEMGzYMgObNmxMYGMjy5ctZvHgxAH/++Se//vorK1eupG3btgA0aNCAzp07s23btiJ34S3zKcXx8fHcvHlTe+RX9uQ5IYQwSVmZ+h9GpFar+e233+jevbtOec+ePbl37x7nzp0DIDo6muTkZHr06KGtY25uTrdu3YiKitIOR0RGRmJra0tAwN+7pDo6OuLn50dUVFSR8ZR5T2XChAl5ysaOHZtvXVn7Swhhskr58VdgYCAJCQnUqVOHPn36MGrUKCwsLIiNjSUjI4P69evr1Hd3dwcgJiYGHx8flEolQJ56bm5uPH78mPj4eGrXro1SqcTV1RUzM7M89Y4cOVJknGWaVBYsWFCWtxdCCOMx4PGXSqXKd+FcOzs77OzsdMpq1qzJ+PHjady4Mebm5kRFRbFs2TJu3brFZ599pn0x/Onrcj/nnlepVFhZWVGhQgWdelWqVAEgMTGR2rVro1KpdMZunmxPn5fQyzSp9O7duyxvL4QQxmNAUlm7di1Lly7NUz5u3DjGjx+vUxYQEKDzKKp169bY2tqyZMkSxowZU/x4S0mZP/4SQohywYDHX0FBQ/P9pfrp3kZBunXrxpIlSzh37pz2MdfTPZ/cz7k9ETs7O9RqNenp6VhbW2vr5fY+7O3ttfXyW8BXpVJp2yqMJBUhhDAGAwbg83vMVVwuLi5YWloSExNDmzZttOVXr14FwNXVFfh7LEWpVOose6VUKqlcuTIODg7aeseOHUOj0aBQKHTay22rMGU++0sIIcqFZ7ig5E8//YRCoaBRo0ZYWVnRokUL9uzZo1Nn165d1KxZE29vbwD8/PywtbVl9+7d2jpZWVns2bOHgIAAbQJp27YtKpWKw4cPa+vFxcURHR2tk7QKIj0VIYQwhlKa/TVs2DD8/f1p0KABCoWCw4cPs2nTJt566y2cnZ2BnBmzAwcOZPbs2QQGBhIdHU1ERARz587VzuKysrJi9OjRhIaGUq1aNe3Lj7GxsSxcuFB7vyZNmtCuXTtmzZrFjBkzsLGxYfHixdpZZ0VRaJ5eK6UcybgfU9YhCBNT0TGg6ErihZSpvl2i61O3fqJ33Ypvzda77qeffkpUVBTx8fFkZmby0ksv0adPH4KCgjA3N9fWi4yM5KuvvkKpVFKrVi2Cg4MZPHhwnvbCw8PZsGED9+/fx93dnalTp9KyZUudOikpKYSEhLB3717UajX+/v7Mnj1bm8QKI0lFvFAkqYiClDipfD+v6Er/r2LfuSW6lymTx19CCGEM5ff3c4NIUhFCCGPIlE26QJKKEEIYRznfe15fklSEEMIYZOdHQJKKEEIYh4ypAJJUhBDCOKSnAkhSEUII45CkAkhSEUIIo9BkZZV1CCZBkooQQhiD9FQASSql7vfo/zF0/PQ85bY2lTn+89ZCr4376y5LVq3j91P/4+HDJGo71OT1DgEMH/QOlSpWKPRaY7oac4PPv17B6bMXsLK0pP1rLZg6fgRV7P7eyOfoiZOEb4hAeT0WVXIy1eyr8IqPF2OGDqD+y/WeWayiaHXrOrLwyw/p1DFnEcEDBw8zecq/uHnzTlmH9nyTKcWAJJVnZuZ7o2jUsIH285Nr9uTncWoaw9/7gMzMTMYPH0Qdh1qcvXiZsNUbuHHzDgs/nlnaIQNw994DhoyfxssuzoR+MgtVSgoLw8IZM/VfrF/+pXaxuiRVMl4ebvTr04Oq9lWIi79H+PrveXfkJLavX45jbYdnEq8oXMWKFdj/8/ekq9MZMuw9NBoN8z6axi/7IvB9tROPH6eWdYjPr2yZ/QWSVJ4Z15dcaNKood71T/3vHDdu3mbFV5/Q2v9VAJq/2oQkVTL//u4HUtPSqFihZL2VsPAN/Lh7P/t+WFtgnTWbtpKZmcXSkA+xs7UBoFaN6gSPncaBqON0btcagO6d29G9czuda30aehD47gj2HTpCcP83SxSrMI7hwwbg6uqCV6M2KJXXAThz5gIXzx9h5IhBLFq8smwDfJ7J4y9A9lMxWRn/v+SDTeVKOuW2NjZkZ2t0psRnZmaxat0WAvuPwLddIO3/MYAvlqwiPV1d4jgOHfmNNi2baRMKQNNXfKjjUItDh48Xeq19lZzHY0X1ysSzE9izCydORGsTCsD16zc5duy//COwS9kFVh5kZel/lGOSVJ6R6R+F0DigB6279WXah58T99fdQuu3bOpLPWcnQpd/i/LaDR4/TuXEydNsiNhB3ze664ypzJgXwsq139G9czuWfTGP4YP6sm3Xz0z/KKREMaelp3M7Lh4315fynHN72QXl9dg85VlZWWRkZHDj5m0+CllCjepV8/RgRNnx8mrA2XOX8pSfO3+Zhk88nhXF8Aw36TJl8virlNnaVCKofx+aveJD5cqVuHhZyap1Wxhw6gwR/15K9ar2+V5nbW3FuuVfMumDT+g1cJS2/M3ArsyaPEb7+eTps+w9EMWns6fQq1snAFo286WKnS0z5n3BxctKPBvkbCOalZWl08PR/P9/3JmZur85WVjk9CxUqhQ0Go1OLyWXnZ0t12Jv5SnvP2IS5y9dAcClriPhX39W4HcUz161avYkJibmKX/4MJGqVYvef1wUQsZUgEKSSn6bu+hDoVCwdm3Bz+hfNA0buNGwgZv2czPfxrz6ig/9R0xkY8SPTBgZlO916elq3p+zgISHiSyYO5U6DjU5c/4S36zZhLm5GXOnjgfgyIk/sLS0oEv713SSQ6vmfgD88edZbVLp1ncod/LpIb3StqfO55+3/hunOsUbWF8w930ePXrMrTt/8e/vfmDkex+wbvnCYrcnxHNDZn8BhSSV33//vVgN5u5zLArm5eFGPWcnzl64XGCdbbt+5r+n/sfuLeG41HUEcsYybG0q8+HnX9P3jR54uruS8DCJjIxMmnXsnW87iUkq7Z+XhnyIWp2h/bz1P3uIPPo7Sz7/l841tWpUA8DWtjIKhQJVckqedlWqZJ0pxbnqv+QCQGNvT15r0ZTX3wpm9frv+de08QV+V/HsPHyYhL193p5j1ar2PHyYVAYRlSPSUwEKSSoXL158lnG8kApLwJeV17GztdEmlFyNGnoAEHMjFk93V+yr2GJtZcXaZV/k206tGtW1f25Q/2Wdc5HHfsfS0kJnqvOTKlaogFMdB65eu5HnnPJ6LE1f8SkwfgA7WxucnRy5eVvefzAV589fxtsr78/bq6E7Fwr5JUcUTVPOx0r0JQP1ZeDshctcj72NTyEDozWqV0WVnELsLd1/kM+cz0n2DjVqANDavynpajUpjx7RqGGDPEetmtXztG2Idq39OXz8vySnPNKWRf95ljt/3aX9ay0KvfZ+wkOuxd7E2alOiWIQxrNz1z78/f14+WUXbVm9enVp1aoZO3ftL8PIygGZ/QXIQH2pm/7h5zg51qZhAzfsbCtz4bKS1eu/p1bN6gx4uxcAd/6Kp1vfoYwKfpfRQwcA8Eb3zqzbvI3R789l5OB3/v/lxyus+PcmvDzc8W3sBUBzv8Z079yOybPnM/id3vh4eaBQKLjzVzxRx//L5NFDecmlbrHjHzLgLXbtO8T46R8yfNA7JKc84qtl39LYy4OObVtp602YOQ+vBm40cHsZm0qVuH7zNuu3bMfC3Jygfn1K8DcojGl1+EbGjA5m2w/fMvdfIWg0Gj76cBo3b95h5ar1ZR3e800efwEGJpXs7Gw2btzIzp07USqVpKamcv78eQDOnz/P999/T1BQEC+//HIRLb043FxfYvcvv7Jp639IS0unevWqdGrbirHDBlHVPme2jUYDWVnZZD8xNcupjgObVoay7NuNfL1qHYmJKmo71OStXt0YObif9k12gM/mTmXj1v+w/ad9rFy3GStLSxzrONC6+atUr1a1RPE71KzBt0s+I+TrVUya9QmWTyzT8mQMTbw9+fngYdZu3kZGRia1a9WgmV9jhg96RwbpTcjjx6l0fr0vC7/8kLVrvkahUHDw0BEmT/kXjx49Luvwnm/y+AsAhUaj384yarWaESNG8Pvvv1OlShWsrKy4d+8eFy5cAEClUtG6dWtGjBjBhAkTSjVofWXcjynrEISJqegYUNYhCBOVqb5dousfze2nd93K8zaX6F6mTO8xlfDwcE6cOMHYsWM5duwYb7/9ts55Ozs7mjVrxpEjR4wepBBCmDxNtv5HOaZ3Utm5cyd+fn6MGzcOMzOzfGcu1a1blzt3ZKaPEOIFlK3R/yjH9E4qt27dokmTJoXWqVKlCklJMtddCPHi0WRm6X0U16NHj2jTpg0eHh6cOXNG59yOHTvo2rUrPj4+9OjRg927d+e5PiMjg4ULF/Laa6/RpEkTBg4cqB3CeNK9e/d47733ePXVV2natCnvv/8+CQkJesWod1KxtrYmOTm50Dp37tzBzs5O3yaFEKL8eAY9laVLl5KVz5TkvXv3Mn36dDp37syqVato2bIlkydPJjIyUqfeggUL2LhxIxMmTGDZsmVYWloSHBxMfHy8tk5mZibDhw/n8uXLfP7553zyySecOnWKMWPGoM8QvN6zvzw9PTl69ChqtRorK6s855OTkzly5Ai+vr76NimEEOVHKY+VXL58mc2bNzNjxgzmzp2rc27x4sV07dqVKVOmANCiRQtiYmJYsmQJbdu2BSA+Pp7Nmzcza9Ys+vbtC0CTJk3o2LEja9euZdq0aQDs27ePixcvsmvXLtzd3QGoVasW/fv3JyoqStteQfTuqbzzzjvExcXx/vvvk5Kiu2yHSqVixowZqFQq+vfvr2+TQghRfpRyT2XevHkMGDCAl156Saf85s2bxMTE0KNHD53ynj17cubMGe1jqyNHjpCVlUX37t21dWxsbGjfvj1RUVHassjISBo0aKBNKAB+fn44OTnl6fnkR++eSs+ePTl69Cjbt2/n4MGDVKmS845Fnz59uHr1Kmq1mgEDBhSZxYQQojzSlOIA/I4dO7hx4wYrVqzg7NmzOudiYnJenahfv75OuZubm/Z8tWrVUCqV1KhRg6pVq+apt2vXLrKzszEzM0OpVGqvfbpe7r0KY9DLjwsWLKBZs2asW7eOS5cuodFoOH/+PO7u7gQHB/Pmm7K7nxDiBWXAALxKpUKlUuUpt7OzyzMunZyczBdffMH06dOpXLlynmtyJ0c9fV3uL/6551UqFba2eReBrVKlChkZGTx+/BgbG5sC69nZ2aFUKov8bgYv09KnTx/69OlDWloaSUlJ2NraUqlSpaIvFEKI8syAnsratWtZunRpnvJx48Yxfrzuit6LFi2iXr16/OMf/yhxiM9Csdf+qlChAhVKuEe6EEKUGwYklaCgIHr3zrtdxdO9jStXrrB582a+/fZbbc/m8ePH2v+bkpKi7ZGoVCpq1qypvTa3h5J73s7OLt8ZvElJSVhaWmo7BwXVU6lU2rYKY3BSefToEfv37+fChQskJydja2tLw4YN6dy5c75dMyGEeBHoueIVkP9jrvzcuHGDzMzMfDdNHDx4MJ6entoeT0xMjM64Su6jKldXVyBnzOXBgwckJibq7KmjVCp56aWXtGv51a9fP993V65evUq7du2KjNmgpLJnzx4+/PBDVCqVzl+gQqFgwYIFfPTRR3Tt2tWQJoUQonwohYF6Pz8/1q1bp1N24cIF7b+33t7eODs74+rqyu7du+ncubO23q5du/Dx8aFatZxN91577TXMzMzYs2ePdpbuo0ePOHjwoM54eNu2bfnxxx9RKpXaJHX69Glu376t10QsvZPK0aNHmTJlCmZmZrzxxhs0b96cGjVqcP/+fU6cOMGuXbuYMmUKdnZ2tGrVqugGhRCiPCmFpFKtWjX8/f3zPeft7Y2PT85GeRMmTGDSpEm4uLjQqlUrDhw4wNGjR1mxYoW2voODA/369ePLL7/EwsICR0dHvv32WyDncVyuLl264OHhwYQJE5g8eTJZWVmEhITg6+tLmzZtioxZ76QSFhaGlZUVGzduxNvbW+dc7969GTBgAAMHDiQsLEySihDihaPJLLuFIrt160ZaWhrffPMN4eHhuLi4sHDhwjw9i5kzZ1KpUiUWLVpEcnIyPj4+rFmzBgeHv7ensLCwYPXq1Xz66adMnToVhUJBu3btmDVrll7bxeu99L2vry9du3ZlwYIFBdaZMWMGP//8M6dOndKnyVInS9+Lp8nS96IgJV36PmlQR73rVll/oET3MmV691SsrKyoVatWoXVq1aqV7xIuQghR3pXmy4/PE72Tyquvvkp0dHShdaKjo2natGmJgxJCiOeOJBXAgLW/3n//fS5dusSXX36pnSed6/Hjx4SEhHDlyhXtgmZCCPFCyTbgKMcKHFOZOXNmnrJbt27xxx9/YGtri5eXF9WrV+fBgwecP3+e5ORkmjZtirOzM/Pnzy/1wPUhYyriaTKmIgpS0jGVh2+307tu1YhfS3QvU1ZgUvH09CxegwpFvi/OlAVJKuJpklREQUqaVBJ667+YbrXtRa/2+7wqcEzlwIHyOztBCCGMrpw/1tJXgUnFycnpWcYhhBDPtVLeo+u5UewFJYUQQjxBkgpQzKSSlZXFw4cPUavV+Z53dHQsUVBCCPG8kZ5KDoOSyqVLl1i4cCEnTpwoMKEoFArOnz9vlOCEEOJ5ocks6whMg95JRalU0q9fPwBatWrFoUOH8PT0pHr16pw/f56HDx/i7+8vvRQhxAtJeio59E4qy5YtIzMzk61bt+Lh4YGnpyedOnVi3LhxPH78mE8++YSoqCg+++yz0oxXCCFMkiSVHHq/Uf/777/Tvn17PDw88pyrVKkS8+bNw87OjsWLFxs1QCGEeC5oFPof5ZjePZWHDx9Sr169vy+0sCA1NVXns7+/P/v37zduhEII8RyQnkoOvZOKvb29zppf9vb2xMXF6dSxtLQkJSXFeNEJIcRzQpNdvnsg+tI7qTg7O3P79t/LGDRq1IijR4/y4MEDqlevzuPHjzlw4AB169YtlUCFEMKUZWdJUgEDxlRat27NiRMntL2Vfv36kZSUxBtvvMGECRMIDAzkzp07vPXWW6UWrBBCmCpNtv5HeaZ3T6Vv3764urqSlpZGpUqVaNeuHTNnziQsLIx9+/ZRsWJFRowYweDBg0szXiGEMEny+CuH3tsJFyT37frq1avrtX/xsySrFIunySrFoiAlXaU4tqn+2wm7/FF+F+wt8dpf5ubm1KhRwxixCCHEc0t6KjlkQUkhhDACGajPUWBSKe7YiEKhYO3atcUOSAghnkfSU8lRYFL5/fffi9WgqY2rCCHEs6Ap52/K66vApHLx4sVnGYcQQjzXyvtUYX3JmIoQQhhBtvRUAANefhRCCFEwjUah92GIffv20b9/f/z9/fHx8aFTp058/vnnJCcn69SLjIykd+/e2jrr16/Pt73w8HA6dOhA48aN6dOnD8ePH89TJyUlhblz5+Lv74+vry+jRo3i1q1besUrSUUIIYwgO0uh92GIpKQkmjVrxscff8zq1asZPHgwP/zwAxMnTtTWOXXqFGPGjKFhw4asWrWKPn36MH/+fL777judtsLDwwkNDWXAgAGsWLGCl156iZEjR+YZ7pgyZQoHDx5kzpw5hIaGcvfuXYKDg3UWES5IiV9+NGXy8qN4mrz8KApS0pcfz9fvoXddL+VPJbrXli1bmDt3LlFRUTg4ODB8+HCSkpKIiIjQ1pkzZw6HDh0iKioKMzMz1Go1rVq1om/fvkybNg3IeXk9MDAQd3d37bYlf/75J3379mXlypW0bdsWgDt37tC5c2c++OADBgwYUGhs0lMRQggjyNYo9D5KqmrVqgBkZGSgVqv57bff6N69u06dnj17cu/ePc6dOwdAdHQ0ycnJ9Ojxd/IzNzenW7duREVFkdu/iIyMxNbWloCAv38Bc3R0xM/Pj6ioqCJjk4F6IYQwAkPGSlQqFSqVKk+5nZ0ddnZ2+V6TlZVFZmYmV65cISwsjA4dOlC3bl2uXr1KRkYG9evX16nv7u4OQExMDD4+PiiVSoA89dzc3Hj8+DHx8fHUrl0bpVKJq6srZmZmeeodOXKkyO8mSUUIIYzAkIGEtWvXsnTp0jzl48aNY/z48fle4+/vrx2cDwgIYOHChUDOmAuQJxnlfs49r1KpsLKyokKFCjr1qlSpAkBiYiK1a9dGpVJha2ub5/52dnbatgojSUUIIYzAkMdaQUFB9O7dO095Qb0UgPXr15OamsqVK1dYvnw5o0aNYs2aNcWKtTSVaVK5c+eOQfUdHR1LKRIhhCiZbAOWaSnsMVdBGjZsCICfnx/e3t68+eab7N+/Hzc3N4A8j9NyP+f2ROzs7FCr1aSnp2Ntba2tl9v7sLe319Z7elff3PZy2ypMgUmlY0f9l3F+kkKh4JdfftGrbocOHQxa1uXChQsGxeLh+aZB9YUQorie5cuPDRs2xMzMjNjYWDp06IClpSUxMTG0adNGW+fq1asAuLq6An+PpSiVSry8vLT1lEollStXxsHBQVvv2LFjaDQanX+fr169qm2rMAUmleLONDbkuvnz58taYUKIcuFZrv116tQpsrOzqVu3LlZWVrRo0YI9e/YQHBysrbNr1y5q1qyJt7c3kNPDsbW1Zffu3dqkkpWVxZ49ewgICND+W9y2bVvCwsI4fPiwNknFxcURHR3NBx98UGRsBSaVgwcPFvsL66tPnz6lfg8hhHgWSqunMmzYMFq0aIG7uzvW1tZcuHCB8PBwPDw86NSpEwBjx45l4MCBzJ49m8DAQKKjo4mIiGDu3LnaWVxWVlaMHj2a0NBQqlWrhpeXFxEREcTGxmoH/QGaNGlCu3btmDVrFjNmzKDG7ggAACAASURBVMDGxobFixdTp04dvf7NLtcvP7rW8C3rEISJiVXdLesQhIkq6cuPvznq/0tyizvb9K67aNEiDhw4oF0mpW7dunTp0oUhQ4ZgY2OjrRcZGclXX32FUqmkVq1aBAcH57uFSXh4OBs2bOD+/fu4u7szdepUWrZsqVMnJSWFkJAQ9u7di1qtxt/fn9mzZ+Ps7FxkvCaVVB48eMCuXbu4du0a6enpOucUCgXz5883qD1JKuJpklREQUqaVI7Wfkvvuq3/2lqie5kyg2d/qdVq/ve//3H37l3UanW+dd544w2DA4mJiaFfv35kZmaSmppK1apVSUpKIisriypVquhkZCGEMDWy8n0Og5LK1q1b+eKLL/J9ExTQzhYoTlIJCQnBx8eHsLAwXnnlFVatWoWHhwc7duxgyZIlhIWFGdymEEI8Kxpk0hEYsPZXVFQUs2fPplatWkyfPh2NRkPHjh2ZNGkSrVq1QqPR0LVrV4MfUeU6e/Ys/fv3x8rKCoDs7GwsLCx46623GDBgAJ9++mmx2hVCiGchW6P/UZ7pnVTWrFmDvb093333nXbamqenJyNHjiQ8PJyPP/6Y/fv36zWQk59Hjx5hb2+PmZkZtra2PHz4UHvOx8eHs2fPFqtdIYR4FrJR6H2UZ3onlfPnz9O+fXudsY0nx/jffvtt/Pz8+Oabb4oVSN26dbl37x4AL7/8Mnv37tWe+/XXX/Ndi0YIIUyFBoXeR3mm95jK48ePqVWrlvaztbU1KSkpOnUaNWrEDz/8UKxAWrVqxbFjx+jWrRvBwcFMnjyZkydPYmFhQUxMDKNGjSpWu0II8SxklfNkoS+9k0rNmjVJSEjQ+Xzt2jWdOsnJyWRlZRUrkClTpmhnk3Xv3p0KFSqwe/du0tLSGDx4MH379i1Wu0II8SzI7K8ceicVNzc3nSTStGlTfvrpJ/744w+aNm3K5cuX2bNnj3YNf0NZWVlpB+khZ12wDh06FKstIYR41iSp5NB7TKVNmzZER0cTHx8PwPDhwzE3N2fQoEG0aNGCXr168ejRI0aPHl2igBISEjh06BDbt28nMTERgPT0dLKz5UcmhDBdMqaSQ+836jMyMkhKSsLOzk7bozh9+jTLly8nNjYWJycngoKCdLagNIRGoyEkJIQNGzaQkZGBQqFg69ateHt7M2zYMPz8/Bg7dqxBbcob9eJp8ka9KEhJ36jfWbu/3nUD//quRPcyZXo//rK0tKRGjRo6Za+88gorVqwwSiArVqxg48aNjB07llatWumMobRv354ff/zR4KQihBDPSnmfKqwvk9n5MSIigrFjx/LPf/4zz2C/i4sLsbGxZRSZEEIUrXhTlMofk0kq8fHxNGnSJN9zlpaWpKamPuOIhBBCf9myNxRgQFLx9PTUa0MthULB+fPnDQ7EwcGBK1eu0KJFizznLl26RN26dQ1uUwghnpVyvvqK3vROKs2aNcu3PDk5mevXr5OWloanp2ex33zv2rUrYWFheHl58corrwA5CeratWt8++238p6KEMKkyfzUHEbZTyUlJYUFCxZw6tQpvvvuO6pUqWJwG2lpaQwdOpRTp07h6OjI7du3cXZ2Ji4uDl9fX8LDw3XeY9GHzP4ST5PZX6IgJZ399Z3jAL3r9r+zsUT3MmVG26QrOzubXr168eqrr/Lhhx8Wq42srCx27tzJkSNHSEhIwN7enoCAALp06cLWrVsJCgoyqD1JKuJpklREQUqaVDY4DtS77sA7G0p0L1NmtIF6MzMz/P392bt3b7GSSkJCAlWrVuWNN97Q7seSmprKd999x+uvv86DBw8MTipCCPGsZMs4PWDAG/X6UKvVBW7gVVD9Tz75BF9fX1q3bo2/vz+bNm0C4Mcff6Rz586EhIRQp04dVq9ebcxQhRDCqLINOMozo/VUlEole/fupV69enpfExYWxoYNG2jVqhVeXl7cunWL+fPno1Qq2bhxIy+99BLz5s2TNcCEECZPZn/l0DupzJw5M9/yrKws4uLiOHXqFFlZWUyfPl3vm+/evZt3332XuXPnasu2bt3K7Nmzad26NcuXLzd4cF4IIcqCPP7KoXdS2b59e6HnXV1dGTZsGG+++abeN4+Li6Nz5846ZV26dGH27NkEBwdLQhFCPDfK+2MtfemdVA4cOJBvuZmZGXZ2dlSuXNngm2dmZua5LvdztWrVDG5PCCHKSpb0VAADkoqTk1OpBBAfH8/Nmze1n3PX/YqPj8fOzk6nrrOzc6nEIIQQJSU9lRwGjal06tSJjh07Fljn0KFD7Nu3jwULFugdwIQJE/Itz29F4gsXLujdrhBCPEullVT27NnDzp07OXfuHElJSTg7O9O/f3/69euHmdnfE3gjIyNZtGgRV69excHBgaCgIAYNGpSnvfDwcDZu3Mj9+/dxc3Nj6tSptGzZUqdOSkoKISEh/Pzzz6jVavz9/Zk9e7Zey2UZNKbi5ORUaFK5ePEiO3bs0DupGJJ8hBDClJXW7K81a9bg6OjItGnTqF69OidOnODTTz/l5s2b2olRp06dYsyYMfTq1Yvp06cTHR3N/PnzsbCwoH//v/d5CQ8PJzQ0lEmTJuHl5UVERAQjR44kIiICT09Pbb0pU6Zw7tw55syZg42NDV9//TXBwcHs3LmTihUrFhqvUVcpVqvVmJub612/d+/exry9EEKUmdKa/fXNN9/ojDG3aNGCx48fs3HjRiZNmoSVlZV23cT58+dr68TFxREWFsY777yDmZkZarWa5cuXM3jwYIYNGwZA8+bNCQwMZPny5SxevBiAP//8k19//ZWVK1fStm1bABo0aEDnzp3Ztm0bAwYUvhyNQS8/FrZKsVqt5o8//sizkZcQQrwISuvlx/wmLTVs2JD09HQSExNRq9X89ttvdO/eXadOz549uXfvHufOnQMgOjqa5ORkevTooa1jbm5Ot27diIqKInfFrsjISGxtbXV28XV0dMTPz4+oqKgi4y20p/L0o661a9eybdu2PPWys7NJSEhArVbTr1+/Im8qhBDlzbPcpOvkyZPY29tTvXp1rl27RkZGBvXr19ep4+7uDkBMTAw+Pj4olUqAPPXc3Nx4/Pgx8fHx1K5dG6VSiaurq854TW69I0eOFBlboUnlybUmFQoFGo2G/NaftLCwoEGDBrRs2ZLRo0cXeVMhhChvDHn8pVKp8l3Sys7OLs+s16edOXOGbdu2MXbsWMzNzUlKStJe+3RbgPa8SqXCysqKChUq6NTLXVU+MTGR2rVro1Kp8t3CxM7OTttWYQpNKgcPHtT+2dPTk6CgIMaNG1dko0II8aIx5LHW2rVrWbp0aZ7ycePGMX78+AKvu3fvHhMmTMDHx4cRI0YUI8rSp/dA/bp160rtXRUhhHjeGTL7KygoKN+JSoX1UpKTkxkxYgQVKlRg+fLlWFpaAn/3NJ7u+eR+zj1vZ2eHWq0mPT0da2trbb3c3oe9vb22XlxcXJ77q1QqvfbK0jupNG/eXN+qQgjxwsk2IK3o85jrSenp6YwePZoHDx6wefNmqlatqj3n4uKCpaUlMTExtGnTRlt+9epVIGcJLfh7LEWpVOLl5aWtp1QqqVy5Mg4ODtp6x44dQ6PR6EzOunr1qratwug9+2vZsmV4e3sTHx+f7/n4+HgaNWrEypUr9W1SCCHKjSwDDkNkZmYyceJELl26xKpVq/I8MbKysqJFixbs2bNHp3zXrl3UrFkTb29vAPz8/LC1tWX37t1/x5yVxZ49ewgICNAmkLZt26JSqTh8+LC2XlxcHNHR0TpJqyB691QOHTpE8+bNtdnsaQ4ODvj7+3PgwAFGjhypb7NCCFEulNYb9fPmzePQoUNMnTqVtLQ0Tp8+rT3n5uaGjY0NY8eOZeDAgcyePZvAwECio6OJiIhg7ty52llcVlZWjB49mtDQUKpVq6Z9+TE2NpaFCxdq22zSpAnt2rVj1qxZzJgxAxsbGxYvXkydOnXo06dPkfHqnVRiY2MJDAwstE79+vX5z3/+o2+TQghRbpTWy4+503i/+OKLPOfWrVuHv78/vr6+LFu2jK+++oodO3ZQq1YtZs6cqfM2PaB96XH9+vXcv38fd3d3Vq5cqfM2PcDChQsJCQnho48+0i7Tsnjx4iLfpgcDkkpaWlqRDVpbW/Po0SN9mxRCiHLDkDEVQzw5C7cwbdu21b4BX5hhw4Zpk0tBbGxsmDdvHvPmzdPr3k/SO6nUrl1bp9uVn9OnTxf4eEwIIcoz2fkxh94D9QEBAfzxxx86gzxP+umnn/jvf/+r10COEEKUN7JHfQ69eyojRoxg586dTJkyhd27dxMQEICDgwPx8fFERUVx8OBBqlSpIoP0QogXUpb0VQADkoqDgwOrV69m4sSJ/PLLLzo7QWo0GpycnFi8eDG1a9culUCFEMKUlfceiL4MWvrex8eHn3/+mUOHDnH69GmSk5OxtbXllVdeoX379to3PIUQ4kVTWgP1zxuD91OxtLSkS5cudOnSJc+57OxsDh48SKdOnYwSnBBCPC8kpeQwaD+Vgty+fZtFixbRrl27QhdDexH5t36VmPun8hynlUXvS/D+rHGsjVjGycuHiLl/ijf7Ff6eUGlx93BlbcQyzlw/ysnLhwj5+kOq2OsuMdEtsBPL1nzJ4VO7OX/zOL/8tp2ps8dT2aZSmcQsCla3riNbNq/kwb0LJNy/SMT3q3B2dizrsJ57MlCfo9g7P2ZlZXHgwAG2bNnC8ePHyc7ORqFQ0KpVK2PGV258OONz/nfqnPZzVlbRizUMHtGPC2cucXDf4TJLKLVq12TTj6uJuXKNsUOnYlfFlhkfvsfqTV/Tt8cQ7VYIw8cO4s6tv/jy0yXE3bmLt48HE6eNosVrTXmrW3C+WyaIZ69ixQrs//l70tXpDBn2HhqNhnkfTeOXfRH4vtqJx49TyzrE55YM1OcwOKncvHmT77//nu3bt/PgwQMAqlatyjvvvMNbb70lKxkXQHn5GqdPnjHomiYvB6DRaKj3snOpJJWJ0/7Jm/3+QRu/HgXWGTluMJaWFgwfMJFkVQoA8X/dY8vOcLp0b8/PP+W8mDViwHskPHiove73YydJfKhi4bKPafFaU44f/q/R4xeGGz5sAK6uLng1aoNSeR2AM2cucPH8EUaOGMSixbJ2X3HJmEoOvR5/ZWZmsmfPHoYMGcLrr7/OqlWrSEpKonPnzmg0Gjp27MjEiRMloRiZvr/dm5ubM3riUPYf38aF2yc4fnYfH8ybjJW1VYlj6Ph6Ww7tP6xNKAD/PR7N7ZtxdOrWTlv2ZELJldszc6hdq8RxCOMI7NmFEyeitQkF4Pr1mxw79l/+EZh3nFToT2PAUZ4V2lO5fv0633//PTt27ODhw4doNBq8vb3p06cPPXv2pEqVKnnWjBH5C/3mU6pWt0eVlMzhQ8cJmfc1d27/ZZS2v/rmUzp2acOKJf8m+vc/qd/gZSbPHENdZ0fGDHm/2O1aV7DGuZ4T32/YkefclYtK3D0KXwbbv/WrACivXCt2DMK4vLwa8J+d+/KUnzt/mbfe7FkGEZUf0lPJUWhS6dq1KwqFgurVqxMcHEyfPn20+x4L/SSrUlgVto4Tx06SkvwIbx8PRr83jK171xLYvh8P7uf9Dd8QzVr4Etj7daaMmcP273cBcDTqBEmJSYR+M5+GjRpw4exlAMzMzHT2R1D8/+ql5ubmOm3mjvdUsbfDzMyMpKS8254mJibxslu9AuNyqF2T96aP5sivv3Hm9PkSfUdhPNWq2ZOYmJin/OHDRKpWLXoDJlGw8j4Ar68ix1QUCgVt2rTh9ddfl4RSDOfPXOL8mUvaz78fO8nvx6PZvm89QSP689WCZSVqv02HVqSnq9mz8xed5HD40G8ANG/5qjap/PrHTuq65J3lcyX+D53PAb7duX0z785v+qpUuSIrNywiKyuLaeP/Vex2hHieaKSnAhSRVCZOnMjWrVvZtm0b27dv5+WXX6Z379706tWLWrXkOXlxnfvfRa4pY2ns613itqrXrIa1tRXnbx7P93zVan//9jliwESdcZZ+g/vQoUsbRg58T+eau3/dA0CVlEx2djZVquTdoc7evgpJiXl7MNYVrFm1cTHO9Zzo/4/h/BV3t1jfS5SOhw+TtNvGPqlqVXsePkwqg4jKD5n9laPQpDJ69GhGjx7N4cOHiYiI4ODBgyxcuJBFixbRunVr3njjjWcVZ7lkjGm2iQlJpKWm8U5g/ktZx/9/ggC4dOGqzrkOXQLIUGcU+HgqLTWNW7F3cPfMO3bi5uHKiWMndcosLCxYtuYLfF7xYvCbo/PcT5S98+cv4+3VIE+5V0N3Lly4XAYRlR/y+CuHXrO/AgIC+Prrr4mMjGTy5Mk4OjoSFRXFlClTUCgUXLhwgbNnz5Z2rOWGzyteuLrV488n3lsprsiDR6lQsQK2tjacOX0+z3H3iaRSHAf2RtK+cwC2tjbasqb+r1DXxZEDeyO1ZQqFgtBvPqXla80YNWiywdOnxbOxc9c+/P39ePllF21ZvXp1adWqGTt37S/DyJ5/2RqN3kd5ptAU89fl48ePs2XLFg4cOEBGRgYKhQIPDw/efvttBgwYYOw4i8W1hm9Zh0DoN59y88Ztzv3vIipVMl4+noyeOIS01DQCO7zLw4REHOvW4dc//sOSL1ex5Mu/3xNo3upVqlevSo1a1fno8xmsW72ZE0dzegd7dv6irbdoxXzadmxN+PIN/Bl9Fk12Nk4ujrTr9Boh8xZzTRmbb2z6vKfiULsmP0Vu4cqlGJYv+hZbOxtm/Os97sbf03mp8eMvPmDAkLdZunAVB/cd1mnjrzvxJvMYLFZlGnGUlUqVKhL9x35S09KY+68QNBoNH304DVubyvi+2olHjx6XdYhlJlN9u0TXD6xX9Fa7uTbc2Faie5myYieVXAkJCWzfvp2IiAiuX7+u7bmYAlNIKqMnDiWwz+s4OdehQsUK3Lv7gMgDR1n0+Tfci78PgJNzHQ6f2s3ikG9YHLJCe+2mH1fRonXTfNt98rspFAqCRvTj7XffwNWtHmq1mluxcUQdOsayr8JJTk7Jtw19kgqAR0M3Zn08Bb9mTcjIyGD/nl+ZP/crEp94Bh8V/VO+kwCAPN+rLL3oSQXA2dmRhV9+SKeObVAoFBw8dITJU/7FjRu3yjq0MlXSpPJuvd561910Y3uJ7mXKSpxUnnTixAkiIiL48ssvjdVkiZhCUhGmRZKKKEhJk0r/evqPMX93I++7X+VFsdf+yo+/vz/+/v7GbFIIIZ4LmTL7CzByUhFCiBeVvKeSQ5KKEEIYgUwpziFJRQghjEC2d8ghSUUIIYxAFpTMYZSdH4UQ4kWXhUbvwxA3btxg7ty59OrVCy8vL3r2zH816cjISHr37o2Pjw+dOnVi/fr1+dYLDw+nQ4cONG7cmD59+nD8eN4lnlJSUpg7dy7+/v74+voyatQobt3Sb8q5JBUhhDCCbDR6H4a4cuUKkZGR1KtXj/r16+db59SpU4wZM4aGDRuyatUq+vTpw/z58/nuu+906oWHhxMaGsqAAQNYsWIFL730EiNHjuTixYs69aZMmcLBgweZM2cOoaGh3L17l+DgYFJTi94Z1KjvqZgaeU9FPE3eUxEFKel7Kt2cu+ldd8/NPXrXzc7Oxuz/t6mYMWMGZ8+eZdeuXTp1hg8fTlJSEhEREdqyOXPmcOjQIaKiojAzM0OtVtOqVSv69u3LtGnTgJxtLgIDA3F3d2fx4sUA/Pnnn/Tt25eVK1fStm1bAO7cuUPnzp354IMPilwxRXoqQghhBNkGHIbITSgFUavV/Pbbb3Tv3l2nvGfPnty7d49z53LWGIyOjiY5OZkePf5eQcPc3Jxu3boRFRWlnWgQGRmJra0tAQEB2nqOjo74+fkRFRVVdLx6fzMhhBAF0hjwP2OKjY0lIyMjz6Ox3P2vYmJiAFAqlQB56rm5ufH48WPi4+O19VxdXfMkMzc3N21bhZHZX0IIYQSGjJWoVCpUqrz7EdnZ2WFnl3f/osIkJSVpr326rSfPq1QqrKysqFChgk69KlVy9lxKTEykdu3aqFQqbG1t840tt63CSFIRQggjyNLo/2Br7dq1LF26NE/5uHHjGD9+vDHDeuYkqQghhBEY8lgrKCiI3r3zrmpsaC8F/u5pPN3zyf2ce97Ozg61Wk16ejrW1tbaerm9j9wdQe3s7IiLy7uduEql0rZVGEkqQghhBIZsvlWcx1wFcXFxwdLSkpiYGNq0aaMtv3o1Z+dVV9ecnVtzx1KUSiVeXl7aekqlksqVK+Pg4KCtd+zYMTQaDQqFQqe93LYKIwP1QghhBBoDDmOysrKiRYsW7NmjO015165d1KxZE29vbwD8/PywtbVl9+7d2jpZWVns2bOHgIAAbQJp27YtKpWKw4f/3mwvLi6O6OhonaRVEOmpCCGEEZTWMi2pqalERuZs3X379m1SUlLYu3cvAD4+Pjg5OTF27FgGDhzI7NmzCQwMJDo6moiICObOnaudxWVlZcXo0aMJDQ2lWrVqeHl5ERERQWxsLAsXLtTer0mTJrRr145Zs2YxY8YMbGxsWLx4MXXq1KFPn6J3t5SXH8ULRV5+FAUp6cuPLZ3a6133+O1Dete9desWHTt2zPfcggULtP/QR0ZG8tVXX6FUKqlVqxbBwcEMHjw4zzXh4eFs2LCB+/fv4+7uztSpU2nZsqVOnZSUFEJCQti7dy9qtRp/f39mz56Ns7NzkfFKUhEvFEkqoiAlTSrNHdvqXff3O5Elupcpk8dfQghhBLJJVw5JKkIIYQTl+KGPQSSpCCGEEch+KjkkqQghhBFITyWHJBUhhDCCLNmlHpCkIoQQRmHIG/XlmSQVIYQwApn9lUOSihBCGIH0VHJIUhFCCCOQnkoOSSpCCGEE0lPJIUlFCCGMwJBNusozSSpCCGEE8vgrhyQVIYQwAo30VABJKkIIYRSyTEsOSSpCCGEEskxLDkkqQghhBNJTySFJRQghjCArW8ZUQJKKEEIYhcz+yiFJRQghjEDGVHJIUhFCCCOQMZUcklSEEMIIpKeSQ5KKEEIYgQzU55CkIoQQRiCPv3JIUhFCCCOQx185JKkIIYQRyNL3OczKOgAhhCgPNAb8z1DXr19n2LBh+Pr60qJFCz7++GNSU1NL4VuUnPRUhBDCCEqrp6JSqRg8eDCOjo4sXryYhIQEFixYQEJCAqGhoaVyz5KQpCKEEEaQXUpL32/evBmVSsWOHTuoVq0aAObm5rz//vuMGTMGd3f3UrlvccnjLyGEMAKNRqP3YYioqChatGihTSgAr7/+OlZWVkRFRRn7a5SY9FSEEMIIDEkWKpUKlUqVp9zOzg47OzudMqVSyZtvvqlTZmVlhYuLCzExMcULthSV66QSc/9UWYcghHhBZKhv6113yZIlLF26NE/5uHHjGD9+vE6ZSqXKk2ggJwElJSUZHmgpK9dJRQghTFFQUBC9e/fOU55f8njeSFIRQohnLL/HXIXVze9RmUqlwtXV1dihlZgM1AshhAmrX78+SqVSp0ytVhMbGytJRQghhGHatGnDb7/9xsOHD7Vl+/fvR61W07Zt2zKMLH8KjSxYI4QQJkulUtGzZ0+cnJwYM2YMDx484LPPPqNly5Ym+fKjJBUhhDBx165d45NPPuHkyZNYW1vTo0cPpk6dSsWKFcs6tDwkqQghhDAaGVMRQghhNJJUhBBCGI3JJpVt27bh4eGR73Hs2DG92ujQoQPvv/9+KUcqnrWn/9to1KgRnTp14quvviI9Pb1U73nr1q1SaV+I8sLkX35cvHgxtWvX1ilzc3Mro2iEKcn9b+PRo0fs37+fFStW8OjRI+bMmVPWoQnxwjL5pNKwYUPq1atX1mEIE/TkfxutW7fmxo0b/PDDD8yaNQszM5PthAtRrj2X/5935MgRRowYwWuvvUaTJk3o2bMn3377LVlZWYVel5WVxZw5c/Dz89M+QktNTeWLL76gQ4cONGrUiA4dOrB8+XKys0tnbwRRery8vEhNTdW+JKbPzzY9PZ358+fTs2dPfH19ad26NaNGjcrzBrMQQj8m31PJysoiMzNT+1mhUHDz5k1atmzJwIEDsba25uzZsyxZsoSEhIQCx1DS0tKYPHkyp0+fZv369Xh7e5OZmcmwYcNQKpWMHj0aDw8PTp8+zbJly0hKSmLGjBnP6msKI7h9+za2trbY29vr/bNVq9U8evSI0aNHU7NmTZKSkti0aRP9+vVj9+7d1KxZs4y/lRDPF5NPKt26ddP57Ofnx3fffaf9rNFoaNq0KRkZGXz77bdMnjw5z6OPpKQkRo0axb1799i8eTMuLi4A7Nq1i5MnT7JhwwaaNWsGQMuWLQEICwtjxIgRVK9evTS/niiB3F84csdU9u3bxwcffIC5uTk7duzQ62dra2vLp59+qtPma6+9RqtWrfjpp58IDg4ui68mxHPL5JNKWFgYDg4O2s+VK1fm7t27LF26lMOHD3P37l2dnsyDBw90fru8e/cu7777LlZWVmzevJkaNWpozx0+fBgnJyd8fX112mjdujWLFi3i9OnTdOzYsZS/oSiup3/hePfddxk4cCBg2M929+7drFmzhmvXrpGcnKyta4obIAlh6kw+qbi7u+sM1GdnZ/P2229z9+5dxo8fj6urK9bW1vzyyy988803eaaUXrp0icTERKZMmaKTUAASEhK4ffs23t7e+d47MTHR+F9IGE3uLxwJCQn8+9//ZtOmTTRp0oQ33nhD75/twYMHmTRpEr1792bcuHFUrVoVhULByJEjUavVz/LrCFEumHxSeVpsbCxnz54lJCSEXr16acsPHTqUb/2AgAA8PT358ssvsba2JigoSHvO3t6eunXrsmjRonyvdXJyMm7wwqie/IWjZcuWBAYGEhISQpcuXfT+2f7000/Uq1eP+1Y0xgAADFFJREFUzz77THsuIyPDJHfUE+J58NwllbS0NAAsLS21ZRkZGezcubPAa4YPH465uTnz589Ho9Fon5MHBASwb98+KlWqRP369Us1blG6rKysmDZtGmPGjGHTpk16/2zT0tIwNzfXKfvxxx+LnEkohMjfc5dUXF1dcXJyIjQ0FDMzMywsLFi7dm2R1w0ZMgQzMzMWLFhAdnY2Q4cOJTAwkG3bthEcHMzQoUPx9PRErVZz8+ZNDh48SFhYmEmuAiry17FjR3x8fFizZg0///yzXj/bgIAAfvnlF+bPn0/79u05c+YMGzZsKBfbugpRFp67pGJlZUVYWBjz5s1j+vTpVKlShTfffBNHR0dmz55d6LVBQUFYWFjw8ccfk5WVxYgRIwgPD2flypVs2bKFW7duUalSJZydnWnXrp1Ob0g8H9577z2GDRvG1q1b9frZ9u3bl7i4OH744Qe2bNmCj48P33zzDePGjSvjbyLE80mWvhdCCGE0z+Ub9UIIIUyTJBUhhBBGI0lFCCGE0UhSEUIIYTSSVIQQQhiNJBUhhBBGI0lFPDMeHh4MGjRIp2zJkiV4eHhw4sSJMorKMIbGO2PGDKNsQzxo0CA8PDxK1EZRjBWreLE9dy8/isI9/Q+PmZkZdnZ2eHh48PbbbxMYGFhGkZUeDw8Pmjdvzvr168s6FCFeeJJUyqncN8IzMzOJiYnhwIEDnDhxgrNnzzJz5swyju5vAwYMoHv37jg6OpZ1KEIII5CkUk6NHz9e5/Px48cZMmQIa9euZdCgQdStW7eMItNVrVo1qlWrVtZhCCGMRMZUXhAtW7bE1dUVjUbDmTNnAN3xgZ07d/L222/j6+tLhw4dtNelpqayYsUKevXqxSuvvIKvry/vvPMOu3btyvc+arWasLAwOnXqpN0XPjQ0tMC9SQobo1AqlcycOVO7x3zLli1599132bRpEwDbtm3TPu77/fff8fDw0B5LlizRaevPP/9kwoQJtG7dmkaNGtG2bVvmzp1LfHx8vnGdPXuWYcOG4evri5+fH8HBwZw6daqIv2X9bdu2jfHjx9OxY0caN26Mn58f/fr148cffyz0OrVaTWhoqPbvpFOnTixdurTAv1+lUsmMGTNo27YtjRo1olWrVkyZMkU2IBOlRnoqL5DcZd4UCoVO+Zo1azh69Cjt27fH399fu/uhSqUiKCiI8+fP4+3tzZtvvkl2djZHjhxhypQpXLlyhUmTJum0/95773HgwAFcXFwYOHAgGRkZ/PDDD1y+fNmgWH/99VcmTpyIWq0mICCAHj16oFKpuHTpEqtXr+bdd9+lYcOGjBs3jqVLl+Lk5ETv3r211zdv3lz7561btzJ37lysrKzo0KEDtWvX5saNG0RERHDw4EG+//57ncdv0dHRDBkyhIyMDDp37ky9evW4cOECgwYNokWLFgZ9j4J8+OGHuLm50axZM2rWrEliYiKRkZFMmzaNa9eu8d577+V73cSJEzlz5gxdu3bFwsKCAwcOsGTJEs6ePcvy5ct1frZRUVGMHz+ezMxM2rdvj4uLC/Hx8ezbt49ff/2VdevWFbiJmRDFphHlSoMGDTQNGjTIU3706FGNh4eHxsPDQ3Pr1i2NRqPRfP3115oGDRpomjRpojl37lyea6ZPn65p0KCBZuXKlTrlaWlpmqFDh2o8PDw058+f15b/5z//0TRo0EDTt29fTVpamrb84cOHmo4dO2oaNGigGThwoE5buTH89ttv2rIHDx5o/Pz8NN7e3poTJ07kiSsuLi7Pd3663VwxMTEa7/9r735DmuriOIB/t2lCzKm1Zm/6Z7LW8h/ORYsUKXFIi0RNIsUi+qu1KOhFCL0TX+XfiVCUkUYh6cxMpMIQrZE6igxrWOgm6TLFaStabrvPC59dum3uWY/jgSd+n1e7v3Puuedc4f62c+71btvGZGRkMFarlVP2/PlzRiaTMcXFxWzM7XYzarWakUqlzOPHjzn1b968yZ7fn/vrj+ccjo+Pc+Jms9mrrsPhYIqKihi5XO7V18LCQkYqlTKZmZmMzWZj49+/f2fy8/MZqVTK6PV6Nm6z2ZiUlBRm+/btzMjICKctk8nEJCUlMdnZ2QH1lZDfQdNff6ja2lrU1taisrISWq0Wx44dA8MwOHz4sNcbLfPz8yGXyzmx2dlZtLe3Iy4uDsePH+eUhYWF4eLFi2AYhvNytNbWVgDA+fPnERYWxsYjIyNRXFwccN/b2tpgt9tx8OBBzi8Oj7Vr1wbc1p07d7CwsIDS0lJER0dzylQqFXbv3o2nT5/CbrcDWPyVMjo6CqVSiYyMDE79wsJCrF+/PuBj++OrnRUrVqCgoABOpxMGg8HnfqdPn0ZERAS7HRYWhgsXLgAAWlpa2HhbWxvm5+eh1WoRGxvLaUMqleLAgQMYHh7G+/fvgzEcQlg0/fWH0ul0ABanukQiERQKBfLy8jivYPZISEjwig0NDcHlcoHH43mtTwCLd5UB4MzNDw8Pg8/nQ6FQeNX3lRyW8urVKwBAWlpawPv8U1v9/f3sWtLPZmZm4HK5MDY2hri4OAwPDwMAlEqlV12BQACFQgGLxbLsfk1MTODatWswGAyYnJxk32jqsdRaj6/zqFAoIBAI8PbtWzbmGfe7d+98/v3GxsYALK65/Jp0CFkOSip/KJPJFHBdsVjsFbPZbAAWk4uvi7HH169f2c9fvnxBRESEz5ebrVmzJuD+eNZ0fv1l8W94xnH9+nW/9b59+8Y5tq9z4i/+O8bHx5GXl4f5+XmkpKRg165dEAqFEAgE+PjxI/R6/ZIL776OHxISgqioKMzMzLAxz7ibm5v99sUzbkKChZIK8Vq4B4Dw8HAAwJEjRwJ+riU8PBxzc3NYWFjwSiyfP38OuD+eY3/69GnZT5ELhUIAgNFoZD8Hcuzp6Wmf5UvFf0dDQwNsNhvKy8uRk5PDKevo6IBer19y3+npaa9nepxOJ2ZnZznj84zj/v37kMlky+4zIYGiNRXiU0JCAvh8PgYHBwPeRy6Xw+12w2g0epX19/cH3E5SUhKAxbuXAsHn8+Fyufy2Feg4PGtLAwMDXmUul8vn2H6X2WwGAGRmZnqV/dN58lVuNBrhcrmwdetWNpaYmMiWEfJfoqRCfFq9ejX27duHN2/eoK6uzudF22KxYHx8nN32fOuuqqqCw+Fg4zabDfX19QEfOzs7G0KhEHfv3vV5cbdarZztyMhIr5hHQUEBQkNDUV5ejtHRUa/yHz9+cBJOcnIyNm3ahIGBATx58oRTt6mpKSjrKZ4bJX5NEL29vbh3757ffevr6zE3N8duOxwOVFRUAAByc3PZeE5ODkQiEXQ6HV6/fu3Vjtvt/t/8vzXy/0LTX2RJly9fhtlsRk1NDdrb25GcnAyxWIypqSl8+PABQ0NDqKiowLp16wAAGo0GnZ2d6O7uhkajwZ49e+B0OtHV1YX4+PiAL8irVq3ClStXoNVqUVRUhLS0NGzZsgV2ux0mkwmTk5Po7u5m66tUKjx8+BCnTp2CXC5HSEgIlEollEolNm/ejLKyMpSWlkKj0SA1NRUbN26E0+nExMQEjEYjoqKi0NXVBWBxKrCsrAxHjx6FVqvlPKdiMBiQmpqK3t7eZZ3XQ4cOobW1FefOnYNarYZEIsHIyAh6e3uRlZWFzs7OJfeNiYnB3r17Oc+pWCwWpKenc27CiIqKQk1NDUpKSpCfnw+VSoXY2FjweDxYrVa8fPkSNpvN73oZIf8GJRWyJKFQiMbGRjQ3N6OjowOPHj2Cw+GAWCzGhg0bcOnSJezcuZOtz+PxUF1djatXr0Kv16OpqQkSiQS5ubkoKSlBfHx8wMdOT09HS0sLe4fUs2fPIBKJEBMTg5MnT3LqlpaWgsfjwWAwoKenB263G2fOnGHv4Nq/fz9kMhkaGhrw4sUL9PX1YeXKlZBIJFCr1cjKyuK0p1AocPv2bVRWVrJTcImJiWhsbERfX9+yk4pMJsOtW7dQVVWFnp4eOJ1OyGQy6HQ6hIeH+00q1dXVqKurw4MHDzA1NYXo6GicPXsWJ06c8FobU6lUaG9vx40bN9DX14fBwUGEhoZCIpFgx44dUKvVyxoHIb7wGObvx6wJIYSQZaI1FUIIIUFDSYUQQkjQUFIhhBASNJRUCCGEBA0lFUIIIUFDSYUQQkjQUFIhhBASNJRUCCGEBA0lFUIIIUFDSYUQQkjQ/AVSPyGJAHdrtwAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ahn38CW-3vj4"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}