MobilenetMasks.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using plaidml.keras.backend backend.\n"
     ]
    }
   ],
   "source": [
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "from keras.applications import MobileNetV2\n",
    "from keras.layers import AveragePooling2D, Dropout, Flatten, Dense, Input, GlobalAveragePooling2D, Conv2D, MaxPooling2D\n",
    "from keras import regularizers\n",
    "from keras.models import Model, Sequential\n",
    "from keras.optimizers import Adam\n",
    "from keras.applications.mobilenet_v2 import preprocess_input\n",
    "from sklearn.metrics import classification_report, confusion_matrix\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pickle\n",
    "import os\n",
    "from keras.callbacks import ModelCheckpoint, EarlyStopping, ReduceLROnPlateau\n",
    "import cv2\n",
    "from keras.models import load_model\n",
    "from PyFunctions import Viz\n",
    "from PyFunctions import Functions as func"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Mobilenet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TRAIN Value Counts\n",
      "1    5000\n",
      "0    5000\n",
      "dtype: int64\n",
      "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n",
      "TEST Value Counts\n",
      "0    509\n",
      "1    483\n",
      "dtype: int64\n",
      "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n",
      "VALIDATION Value Counts\n",
      "1    400\n",
      "0    400\n",
      "dtype: int64\n",
      "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n"
     ]
    }
   ],
   "source": [
    "dim = (224,224)\n",
    "\n",
    "x_train, x_test, y_train, y_test, x_val, y_val = func.get_mask_splits(dim)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_mobilenet(dim):\n",
    "    model = Sequential()\n",
    "    optimizer = Adam(lr = .0005)\n",
    "    baseModel = MobileNetV2(weights=\"imagenet\", include_top=False,\n",
    "        input_tensor=Input(shape=dim))\n",
    "    \n",
    "    model.add(baseModel)\n",
    "    model.add(AveragePooling2D(pool_size=(7, 7)))\n",
    "    model.add(Flatten(name=\"flatten\"))\n",
    "    model.add(Dense(256, activation=\"relu\"))\n",
    "    model.add(Dropout(0.6))\n",
    "    model.add(Dense(2, activation=\"sigmoid\", name = 'Output'))\n",
    "\n",
    "    \n",
    "    for layer in baseModel.layers:\n",
    "        layer.trainable = False\n",
    "        \n",
    "    model.compile(loss = 'binary_crossentropy', optimizer = optimizer, metrics = ['accuracy'])\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:plaidml:Opening device \"opencl_amd_ellesmere.0\"\n",
      "F:\\ProgramFiles\\conda\\envs\\FaceMaskEmotionDetection\\lib\\site-packages\\keras_applications\\mobilenet_v2.py:294: UserWarning: `input_shape` is undefined or non-square, or `rows` is not in [96, 128, 160, 192, 224]. Weights for input shape (224, 224) will be loaded as the default.\n",
      "  warnings.warn('`input_shape` is undefined or non-square, '\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/2000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:plaidml:Analyzing Ops: 1398 of 2052 operations complete\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "157/157 [==============================] - 133s 848ms/step - loss: 0.1937 - acc: 0.9206 - val_loss: 0.3634 - val_acc: 0.8478\n",
      "\n",
      "Epoch 00001: val_loss improved from inf to 0.36340, saving model to ModelWeights/Mobilenet_Masks.h5\n",
      "Epoch 2/2000\n",
      "157/157 [==============================] - 86s 550ms/step - loss: 0.0947 - acc: 0.9663 - val_loss: 0.3759 - val_acc: 0.8584\n",
      "\n",
      "Epoch 00002: val_loss did not improve from 0.36340\n",
      "Epoch 3/2000\n",
      "157/157 [==============================] - 90s 576ms/step - loss: 0.0832 - acc: 0.9703 - val_loss: 0.3132 - val_acc: 0.8790\n",
      "\n",
      "Epoch 00003: val_loss improved from 0.36340 to 0.31321, saving model to ModelWeights/Mobilenet_Masks.h5\n",
      "Epoch 4/2000\n",
      "157/157 [==============================] - 90s 574ms/step - loss: 0.0695 - acc: 0.9755 - val_loss: 0.1480 - val_acc: 0.9425\n",
      "\n",
      "Epoch 00004: val_loss improved from 0.31321 to 0.14802, saving model to ModelWeights/Mobilenet_Masks.h5\n",
      "Epoch 5/2000\n",
      "157/157 [==============================] - 90s 573ms/step - loss: 0.0650 - acc: 0.9771 - val_loss: 0.1396 - val_acc: 0.9461\n",
      "\n",
      "Epoch 00005: val_loss improved from 0.14802 to 0.13957, saving model to ModelWeights/Mobilenet_Masks.h5\n",
      "Epoch 6/2000\n",
      "157/157 [==============================] - 90s 573ms/step - loss: 0.0641 - acc: 0.9779 - val_loss: 0.2344 - val_acc: 0.9047\n",
      "\n",
      "Epoch 00006: val_loss did not improve from 0.13957\n",
      "Epoch 7/2000\n",
      "157/157 [==============================] - 93s 592ms/step - loss: 0.0622 - acc: 0.9785 - val_loss: 0.1219 - val_acc: 0.9395\n",
      "\n",
      "Epoch 00007: val_loss improved from 0.13957 to 0.12195, saving model to ModelWeights/Mobilenet_Masks.h5\n",
      "Epoch 8/2000\n",
      "157/157 [==============================] - 91s 578ms/step - loss: 0.0565 - acc: 0.9798 - val_loss: 0.0544 - val_acc: 0.9798\n",
      "\n",
      "Epoch 00008: val_loss improved from 0.12195 to 0.05442, saving model to ModelWeights/Mobilenet_Masks.h5\n",
      "Epoch 9/2000\n",
      "157/157 [==============================] - 90s 574ms/step - loss: 0.0582 - acc: 0.9790 - val_loss: 0.0747 - val_acc: 0.9677\n",
      "\n",
      "Epoch 00009: val_loss did not improve from 0.05442\n",
      "Epoch 10/2000\n",
      "157/157 [==============================] - 91s 580ms/step - loss: 0.0571 - acc: 0.9801 - val_loss: 0.0426 - val_acc: 0.9869\n",
      "\n",
      "Epoch 00010: val_loss improved from 0.05442 to 0.04260, saving model to ModelWeights/Mobilenet_Masks.h5\n",
      "Epoch 11/2000\n",
      "157/157 [==============================] - 91s 580ms/step - loss: 0.0512 - acc: 0.9801 - val_loss: 0.0371 - val_acc: 0.9894\n",
      "\n",
      "Epoch 00011: val_loss improved from 0.04260 to 0.03707, saving model to ModelWeights/Mobilenet_Masks.h5\n",
      "Epoch 12/2000\n",
      "157/157 [==============================] - 91s 581ms/step - loss: 0.0520 - acc: 0.9802 - val_loss: 0.0797 - val_acc: 0.9662\n",
      "\n",
      "Epoch 00012: val_loss did not improve from 0.03707\n",
      "Epoch 13/2000\n",
      "157/157 [==============================] - 91s 577ms/step - loss: 0.0468 - acc: 0.9832 - val_loss: 0.0659 - val_acc: 0.9698\n",
      "\n",
      "Epoch 00013: val_loss did not improve from 0.03707\n",
      "Epoch 14/2000\n",
      "157/157 [==============================] - 90s 575ms/step - loss: 0.0448 - acc: 0.9838 - val_loss: 0.0643 - val_acc: 0.9728\n",
      "\n",
      "Epoch 00014: val_loss did not improve from 0.03707\n",
      "Epoch 15/2000\n",
      "157/157 [==============================] - 90s 576ms/step - loss: 0.0494 - acc: 0.9823 - val_loss: 0.0421 - val_acc: 0.9849\n",
      "\n",
      "Epoch 00015: val_loss did not improve from 0.03707\n",
      "Epoch 16/2000\n",
      "157/157 [==============================] - 91s 579ms/step - loss: 0.0518 - acc: 0.9815 - val_loss: 0.0518 - val_acc: 0.9824\n",
      "\n",
      "Epoch 00016: val_loss did not improve from 0.03707\n",
      "Epoch 00016: early stopping\n"
     ]
    }
   ],
   "source": [
    "early_stopping = EarlyStopping(monitor='val_loss', verbose = 1, patience=5, min_delta = .00075)\n",
    "model_checkpoint = ModelCheckpoint(f'ModelWeights/Mobilenet_Masks.h5', verbose = 1, save_best_only=True,\n",
    "                                  monitor = 'val_loss')\n",
    "lr_plat = ReduceLROnPlateau(patience = 5, mode = 'min')\n",
    "epochs = 2000\n",
    "batch_size = 64\n",
    "    \n",
    "\n",
    "dim = (x_train.shape[1], x_train.shape[2], x_train.shape[3])\n",
    "mobilenet = get_mobilenet(dim =dim)\n",
    "\n",
    "    \n",
    "augmentation =ImageDataGenerator(rotation_range = 20, width_shift_range = .2, height_shift_range = .2, \n",
    "                                                       horizontal_flip = True, shear_range = .15, \n",
    "                                 fill_mode = 'nearest', zoom_range = .15)\n",
    "augmentation.fit(x_train)\n",
    "mobilenet_history = mobilenet.fit_generator(augmentation.flow(x_train, y_train, batch_size = batch_size),\n",
    "            epochs = epochs, \n",
    "     callbacks = [early_stopping, model_checkpoint, lr_plat], validation_data = (x_test, y_test), verbose= 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Viz.plot_loss_accuracy(mobilenet_history, 'dark_background', path = 'Images/Mobilenet_Loss_Acc.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:plaidml:Opening device \"opencl_amd_ellesmere.0\"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "F1 Scores Test\n",
      "~~~~~~~~~~~~~~~~~~~~~~\n",
      "No Mask(0): 0.9838909541511772\n",
      "Mask(1): 0.9836065573770492\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjgAAAFgCAYAAAC2QAPxAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAABOhElEQVR4nO3deVhU5fs/8PeAIIuACqIICC6ouCWSu6lF7uaeWlZaflzS1Ez7aLaoaYt7ZpZ9yS0/GpqmImquuIuMMMwMm0KhIuIaKuaK3L8/+HlihGGQgMHx/bqu+7o4M895zn3OGZu7c545jwqAgIiIiMiCWJk7ASIiIqLixgKHiIiILA4LHCIiIrI4LHCIiIjI4rDAISIiIovDAoeIiIgsDgscIiIisjgscIio0Nzd3XHw4EHcvHkT8+fPN3c6RqWkpCAoKKjY+hMR1K5du9j6M8Xf3x9qtbpUtmVra4uEhAS4ubmVyvaISgsLHKJ/KSUlBbdv30ZmZibS09OxcuVKODo6GrRp3bo19u3bh5s3b+L69esIDQ2Fv7+/QRsnJycsWrQIZ8+eRWZmJpKTk7Fo0SK4uroqbTw8PJCammqwXnh4OP766y/Y2trmeX348OEGr3Xo0CHP+uPGjYNer8etW7eQmpqKDRs2oFGjRvnu68iRI3H16lU4Oztj8uTJhTtABRg6dCiysrKQmZlpEB4eHv+67ydRrVo1/PTTT7hw4QJu3ryJhIQEzJgxAw4ODqWaxyOzZs1SCsjcx+Xhw4fKZy0zMxOvv/76E/f9+Ofi/v37WLFiBaZOnVps+ROVBSxwiIrBK6+8AicnJzRt2hQBAQH46KOPlPdatWqF3bt3Y+vWrahevTpq1qwJrVaLo0ePombNmgAAGxsb7Nu3Dw0bNkTXrl3h7OyM1q1b49q1a2jRooXSV/fu3fH7778ryz4+PnjhhRcgIujVq9cT57148WJMmDAB48ePR+XKlVG3bl1s2bIFPXr0yLe9j48P4uPjn3g7AGBtbZ3v68ePH4eTk5NBpKenF2kbRVGpUiUcP34c9vb2aN26NZydndGpUydUrFixVK/aPFKtWjW8+OKL2LJlCwAYHJdz584pnzUnJyesW7euWLa5bt06DB06NE+RTPS0EwaDUfRISUmRoKAgZXnOnDkSFhamLB86dEiWLl2aZ70dO3bI6tWrBYAMHz5cLl68KI6OjgVua9OmTdK3b19l+dNPP5UjR47IggULZNu2bQZtw8PDZfjw4QavdejQQVJTUwWA1KlTR7KysqR58+aF2s+VK1fK/fv35d69e5KZmSlBQUFia2srixYtkrS0NElLS5NFixaJra2twbb++9//Snp6uvz88895+hw6dKgcPnzY6DanTJkiycnJcvPmTYmLi5M+ffoYvP+f//xH4uPjlfcDAgKUczJp0iTRarVy/fp1CQkJkfLly+e7jVmzZolOpxOVSmU0DxGR2rVrCwDp3r27REdHy40bN+TcuXMyffp0pV358uVlzZo1cvXqVcnIyJDIyEhxd3dX9vWPP/6Qmzdvyp9//imvv/56vtt68803Zc+ePSY/ayqVSjk+V69elfXr10ulSpUKzGP27NmSlZUld+7ckczMTFmyZInS9+nTp6V9+/Zm//fEYBRjmD0BBuOpjtxfOp6enqLT6eSbb74RAGJvby9ZWVnSsWPHPOsNGzZMLly4IADkl19+kVWrVhW4nXLlysmVK1ekQoUKymtJSUny7rvvSrNmzeT+/fvKlylgusAZNWqUnDlz5on2deXKlTJr1ixleebMmXL8+HGpUqWKuLm5ydGjR+Xzzz9XtvXgwQP5+uuvxdbWVuzs7PL0Z6rAGTBggHh4eIhKpZKBAwfKrVu3pFq1asp758+fl+eff14ASO3ataVGjRrKOTlx4oR4eHhIpUqVJD4+XkaNGpXvNo4fPy4zZswocL9zFzgdOnSQRo0aiUqlksaNG8vFixeld+/eAkBGjhwpoaGhYm9vL1ZWVtKsWTNxcnISBwcHuXHjhtStW1cASLVq1aRBgwb5bmvu3Lny3XffmfysjR8/Xo4fPy6enp5ia2sry5Ytk3Xr1hWYh7HPBQDZunWrjBs3zuz/nhiM4greoiIqBlu2bMHNmzdx/vx5XL58GdOnTwcAVK5cGdbW1vnecklPT1cGdrq6upq8LdO+fXtotVrcunULANC2bVv4+Phgw4YNiI6Oxh9//PFEYzIKs01ThgwZgs8//xxXrlzB1atXMXPmTLz55pvK+9nZ2Zg+fTru37+Pu3fv5ttHq1atkJGRoURycrLy3saNG5Geng4RwYYNG5CUlKTcsvvPf/6DuXPn4uTJkwCAP/74A+fOnVPW/fbbb5Geno6MjAxs27YNTZs2LZbjcPDgQcTGxkJEoNfr8csvv6BDhw4AgAcPHsDV1RV16tRBdnY2oqOjkZmZqRyLRo0awc7ODhcvXjR6q69ixYrKOgUZPXo0Pv74Y6SlpeH+/fuYMWMGBgwYAGtr6wLzMCYzMxMVK1Ys9HEgKutY4BAVgz59+sDZ2RkdOnRA/fr1lcIlIyMDDx8+zHfQrIeHB65evQoAuHbtmsmBtd27d8eOHTuU5aFDh2L37t24du0agH/GUTySlZUFGxsbgz5sbGzw4MGDQm/TlOrVq+Ps2bPK8tmzZ1G9enVl+cqVK7h3716BfURERKBSpUpK1KlTR3nvzTffhEajUYqfRo0aKcfW29sbf/zxh9F+L168qPx9+/ZtVKhQId92T3ocWrRogf379+Py5cu4fv06Ro8ereS0Zs0a7Nq1CyEhIUhLS8OcOXNQrlw53L59G4MGDcLo0aORnp6OsLAw1KtXL9/+MzIy4OTkZDIPHx8fbN68WTk2CQkJePjwIapWrWo0j4I4OTnh+vXrhT4ORGUdCxyiYnTo0CGsWrVK+QXM7du3cfz4cbz66qt52g4cOBD79u0DAOzduxddunQp8Fc7uQscOzs7DBw4EB06dEB6ejrS09MxceJENG3aFE2aNAEAnDt3Dr6+vgZ91KxZUylI9u3bBy8vLwQGBhZ5fy9cuAAfHx9luUaNGrhw4YKyLCJF7rtGjRoIDg7Ge++9B1dXV1SqVAmxsbFQqVQAgNTU1GIZBLx371707dtX6deUdevWITQ0FN7e3qhYsSKWLVumrJuVlYXPP/8cDRs2RJs2bdCzZ0+89dZbAIDdu3ejc+fO8PDwQGJiIoKDg/PtX6fToW7duibzSE1NRbdu3QyKQ3t7e1y4cKHAPIydE39/f2i12kIdA6KnAQscomL2zTffoFOnTkqhMXXqVAwdOhTjxo1DhQoVULFiRcyaNQutW7fGzJkzAeT8n39qaio2bdqEevXqQaVSoXLlyvjoo4/QrVs3+Pr6onz58khMTASQc8Xo4cOHaNCgAZo2bYqmTZvC398fhw4dUr7I1q9fj7fffhvNmzcHAPj5+WHixIkICQkBACQnJ+P7779XbrHY2NigfPnyGDRoEKZMmVKoff3ll1/wySefwM3NDa6urvjss8/wv//9r1iOo6OjI0QEV65cAQAMGzbM4OfrP/30EyZPnoxmzZoBAGrXro0aNWo88XYWLlwIZ2dnrF69Wlm/evXqWLBgARo3bpynvZOTE/766y/cu3cPzZs3N7gt2LFjRzRq1AhWVla4efMmHjx4gOzsbLi7u6NXr15wcHDAvXv3cOvWLWRnZ+ebz549e9CsWTOUL1++wLyXLVuGL774QsnZzc1N+SWdsTwA4NKlS6hVq5ZBX9WrV0flypURERFRyKNG9HQw+0AgBuNpjsd/RQVAvv/+e9m4caOy3LZtWwkPD5fMzEy5ceOGhIWFScOGDQ3WcXZ2lkWLFsm5c+ckMzNTkpOTZcGCBVK5cmUZO3aswS9edu7cKfPnz8+Ty6uvvirp6elibW0tAOTtt9+W2NhYuXHjhiQlJcmUKVPy/Fpo/PjxEhsbK3///becP39eQkJCjA6AfXyQcfny5WXx4sVy4cIFuXDhgixevFj5tVLuAc3GYujQoZKVlSWZmZkG8Wjg8OzZs+XatWty5coVWbBggRw4cMBggOyoUaMkMTFRMjMzRa/XS9OmTfM9J9OnT5c1a9YYzcPDw0OWL18u6enpcvPmTUlISJDPPvtM7O3tBTAcZNy/f385c+aM3Lx5U7Zt2yZLlixR+h48eLAkJibKrVu35OLFi7J48WKxtraWatWqyYEDB+T69euSkZEh4eHh4u/vbzSfDRs2yMCBAwv8rKlUKpk4caIkJibKzZs3JTk5Wb744osC8wAgrVq1klOnTslff/0lixcvFgAyefJkWbBggdn/LTEYxRxmT4DBYJiI7du3S7du3cyeB6N0wt/fXyIjI0tlW7a2tpKQkCBVqlQx+34zGMUZqv//BxGVYR9++CGWLFli9JdIRERkiAUOERERWRwOMiYiIiKLU/CDEcqgy5cvGzx3g4iIiCyXj48P3N3dn3i9p67AOXv2rPKzVyIiIrJsarW6SOvxFhURERFZHBY4REREZHFY4BAREZHFYYFDREREFocFDhEREVkcFjhERERkcVjgEBERkcVhgUNEREQWhwUOERERWZwSK3CWL1+OS5cuQa/XG22zePFiJCUlQavVIiAgoKRSISIiomdMiU3VsGrVKnz33Xf4+eef832/W7du8PPzg5+fH1q2bIkffvgBrVq1Kql0HuMJwA5AeQA2ABIB3MunnROA2rmWMwH8YaRPXwAVcy2fAXA9n3bWABrnWn4IwFgR6A6geq7lywAuGGnrj5z9eSQBT/8+9QPgkGv5NwC382lXHcBLuZYvANhvpM8OALxzLR8EkJpPOzsAA3It3wPwq5E+GwN4LtdyLIAYI225T//gPnGf8sN94j49UtA+maYCIEVe2wQfHx+EhYWhcePGed5btmwZDhw4gJCQEABAYmIiOnbsiIsXLxbYp1qtLmAuKhuEDbdBD//8DhwRERE9LfTpQGMP4GSHgr73jTPbGBxPT0+kpv5TQZ4/fx6enp75th0xYgTUajXUajXc3NzyaVEOwCsIG+7D4oaIiOgplnAJ6LMSaLIA2JdU9H6eitnEg4ODERwcDCD/WUXDhmehh/82ZXm7Ywf0HH2gtNKjkrB9O9BzHwCXXC8uBHAzn8Z1ALyRazkJwFojHQ9Czu28R0KQc4vycY4APsy1/DeAeUb6bA/Dy6qHYPyy6kRwnx7hPnGf8sN9enb3yReAK4Bo5NxcssbLPz6E+j9GujXBbAVOWloavL3/uSfo5eWFtLS0J+skLAzo0QM9FqiUl7bX7I6e/bYXvo/tUUDPDsg5kY+rAqBtruUrAI4a6ag5csb2PKIGkN/+2ADokWv5AQBj+foBaJhrOQlAnJG2nZDz4XxkD57+ffKB4UfU2NW5CwDWFKIdkHPv+WSuZWO3RO8+1ufDAvrUw/C4ZBTQ9jdwn0y1A7hP3Kd/cJ/yZ4n7dOb/BwA4I2dM0OUC+jVNSip8fHxEr9fn+1737t1lx44dAkBatmwpJ06cKFSfarU65++wMIGIhG3qLjIfIvMhEPknwvaX2H4xGAwGg8EonVC+958wSuwKzrp169CxY0e4ubkhNTUV06dPh42NDQDgxx9/xI4dO9C9e3ckJyfj9u3bePvttwvX8fPPAyLKYo+UHQCA7QkAVA7IGfWdXbw7Q0RERE+VEv0VVUlQi0AZS719OySxJwBANdlsKREREVEJKfjX08Y9nU8yVo0CVI5Az57mzoSIiIjKoKezwEF1hA2vAplv7jyIiIioLHoqfiaeW52MJMj8GQavbU8wTy5ERERUNj11BU7FezeUv7cn2KHn8rtmzIaIiIjKoqeuwHlENbkZ8n9QERERET3rntoCJ+dJh0RERER5PaWDjImIiIiMY4FDREREFocFDhEREVkcFjhERERkcZ7iAkdlugkRERE9k57iAmekuRMgIiKiMuopLnDumzsBIiIiKqOe4gLnnrkTICIiojKKBQ4RERFZnKe4wNlk7gSIiIiojHqKCxwiIiKi/LHAISIiIovDAoeIiIgsDgscIiIisjgscIiIiMjiPMUFjrO5EyAiIqIy6ikucCqaOwEiIiIqo57iAocP+iMiIqL8scAhIiIii/MUFzh3zJ0AERERlVFPcYHDKzhERESUv6e4wCEiIiLKHwscIiIisjgscIiIiMjisMAhIiIii8MCh4iIiCwOCxwiIiKyOCxwiIiIyOKwwCEiIiKLwwKHiIiILA4LHCIiIrI4LHCIiIjI4rDAISIiIovDAoeIiIgsDgscIiIisjgscIiIiMjisMAhIiIii8MCh4iIiCwOCxwiIiKyOCxwiIiIyOKwwCEiIiKLU6IFTpcuXZCYmIikpCRMmTIlz/ve3t7Yv38/oqOjodVq0a1bt5JMh4iIiJ4hUhJhZWUlycnJUrNmTbGxsZGYmBjx9/c3aPPjjz/K6NGjBYD4+/tLSkqKyX4lXS0yv2RyZjAYDAaDUbZCrVYXab0Su4LTokULJCcnIyUlBQ8ePEBISAh69+5t0EZE4OzsDABwcXHBhQsXSiodIiIieoaUK6mOPT09kZqaqiyfP38eLVu2NGgzY8YM7N69G+PGjYOjoyNefvnlfPsaMWIERo4cWVKpEhERkYUx6yDj1157DatWrYK3tze6d++ONWvWQKVS5WkXHByM5s2bo3nz5mbIkoiIiJ42JVbgpKWlwdvbW1n28vJCWlqaQZvhw4djw4YNAICIiAjY2dnBzc3NZN/bE4o3VyIiIrIsJVbgqNVq+Pn5wdfXFzY2Nhg8eDBCQ0MN2pw7dw5BQUEAgPr168POzg5Xrlwx2XfP5SWSMhEREVmIEitwHj58iPfeew+7du1CQkICNmzYgPj4eMycOROvvPIKAGDSpEkYMWIEYmJi8Msvv2DYsGEllQ4RERE9Q1TI+TnVU0PS1VB5cCwOERHRs0CtVhdpDC6fZExEREQWhwUOERERWRwWOERERGRxWOAQERGRxWGBQ0RERBaHBQ4RERFZHBY4REREZHFY4BAREZHFYYFDREREFocFDhEREVkcFjhERERkcVjgEBERkcVhgUNEREQWhwUOERERWRwWOERERGRxWOAQERGRxWGBQ0RERBaHBQ4RERFZHBY4REREZHFY4BAREZHFYYFDREREFocFDhEREVkcFjhERERkcVjgEBERkcVhgUNEREQWhwUOERERWRwWOERERGRxCl3g2Nvbl2QeRERERMXGZIHTunVrxMXFITExEQDQpEkTLF26tMQTIyIiIioqkwXOokWL0KVLF1y7dg0AoNPp0L59+xJPjIiIiKioCnWL6vz58wbLDx8+LJFkiIiIiIpDOVMNUlNT0bp1a4gIypUrhwkTJiAhIaE0ciMiIiIqEpNXcEaPHo2xY8fC09MTaWlpaNq0KcaMGVMauREREREVickrOPXq1cMbb7xh8FqbNm1w7NixEkuKiIiI6N8weQVnyZIlhXqNiIiIqKwwegWnVatWaNOmDapUqYKJEycqrzs7O8Pa2rpUkiMiIiIqCqMFjq2tLSpUqIBy5crByclJef3mzZsYMGBAqSRHREREVBQqAFJQgxo1auDcuXOllI5pkq6GyqO5udMgIiKiUqBWq9G8+ZN/75scZHz79m3MnTsXDRs2hJ2dnfJ6UFDQE2+MiIiIqDSYHGS8du1aJCYmombNmpg5cybOnDkDtVpdGrkRERERFYnJAsfV1RUrVqzAgwcPcOjQIQwfPhwvvfRSaeRGREREVCQmb1E9ePAAAJCeno7u3bvjwoULqFy5coknRkRERFRUJguc2bNnw9nZGZMmTcKSJUvg7OyM999/vxRSIyIiIioakwXO9u3bAQBxcXHKrak2bdqUbFZERERE/4LRAsfKygoDBw6Ep6cnfv/9d8TFxaFHjx6YNm0a7O3t0axZs9LMk4iIiKjQjBY4y5cvh7e3NyIjI/Htt9/iwoULeP755zF16lRs3bq1NHMkIiIiemKSX+j1elGpVAJAypcvLxkZGVK5cuV82xqLLl26SGJioiQlJcmUKVPybfPqq69KXFycxMbGytq1a032KenqJ8qBwWAwGAzG0xtqdZG/9/N/IyoqqsBlU2FlZSXJyclSs2ZNsbGxkZiYGPH39zdoU6dOHYmOjpaKFSsKAKlSpYrJflngMBgMBoPx7ERRCxyjt6jq168PrVYLAFCpVKhduza0Wi1UKhVEBM8995yxVQEALVq0QHJyMlJSUgAAISEh6N27NxISEpQ2I0aMwNKlS3H9+nUAwJUrVwrsk4iIiKgwjBY4/v7+/6pjT09PpKamKsvnz59Hy5YtDdrUrVsXAHDkyBFYW1tjxowZ2LVrV56+RowYgZEjR/6rfIiIiOjZYbTAKY0JNsuVKwc/Pz907NgRXl5eOHToEBo3bowbN24YtAsODkZwcDAAQNI5TQQREREVzORUDUWVlpYGb29vZdnLywtpaWkGbc6fP4/Q0FBkZWXhzJkzOH36NPz8/EoqJSIiInpGlFiBo1ar4efnB19fX9jY2GDw4MEIDQ01aLNlyxZ07NgRQM6cV3Xr1sWff/5ZUikRERHRM6JQBY6dnZ0yXqawHj58iPfeew+7du1CQkICNmzYgPj4eMycOROvvPIKAGDXrl24du0a4uLiEB4ejg8//BB//fXXk+8FERER0WMK/JlVz549JTExUf78808BIM8995xs3brVbD8X48/EGQwGg8F4dqKoPxM3eQVnxowZaNGihfJTbq1Wi5o1a5pajYiIiMhsTBY4Dx48wM2bNw1eE5ESS4iIiIjo3zJZ4MTFxeG1116DtbU16tSpg2+//RbHjh0rjdyIiIiIisRkgTNu3Dg0bNgQ9+7dw7p163Djxg28//77pZAaERERUdGokDMYx6iAgABoNJpSSsc0SVdD5dHc3GkQERFRKVCr1Wje/Mm/901ewVmwYAHi4+Px+eefo2HDhkVKjoiIiKg0mSxwXnrpJbz44ou4cuUKfvzxR+h0Onz88celkRsRERFRkRTqQX+XLl3CkiVLMHr0aMTExOCzzz4r6byIiIiIisxkgVO/fn1Mnz4dOp0OS5YswbFjx+Dl5VUauREREREVidHZxB9ZsWIF1q9fjy5duiA9Pb00ciIiIiL6V0wWOG3atCmNPIiIiIiKjdECZ/369Rg0aBB0Op3Bk4tVKhVEBM8991ypJEhERET0pIwWOBMmTAAA9OzZs9SSISIiIioORgcZX7x4EQAwZswYnDt3ziDGjBlTagkSERERPSmTv6Lq1KlTnte6detWIskQERERFQejt6hGjx6NMWPGoFatWtBqtcrrTk5OOHr0aKkkR0RERFQURgucdevWYefOnfjqq68wdepU5fXMzExkZGSUSnJERERERWG0wBERnD17FmPHjs3zXqVKlVjkEBERUZlV4BWcV155BVFRURARqFQq5T0RQe3atUslQSIiIqInZbTAeeWVVwAAtWrVKrVkiIiIiIqDyV9RtWnTBg4ODgCAIUOGYMGCBfD29i7xxIiIiIiKymSB88MPP+D27dto0qQJJk2ahD/++ANr1qwpjdyIiIiIisRkgZOVlQUA6N27N7777jt8//33cHJyKvHEiIiIiIrK5GSbmZmZmDp1Kt5880288MILUKlUsLGxKY3ciIiIiIrE5BWcQYMG4d69e3jnnXdw6dIleHl5Yd68eaWRGxEREVGRmCxwLl26hLVr18LFxQU9evTA3bt3OQaHiIiIyjSTBc6rr76KyMhIvPrqqxg4cCBOnDiB/v37l0ZuREREREVicgzOxx9/jObNm+PKlSsAADc3N+zduxebNm0q8eSIiIiIisLkFRwrKyuluAGAa9euwcrK5GpEREREZmPyCs7vv/+O33//Hb/88guAnEHHO3bsKPHEiIiIiIpKBUBMNerbty/atWsHADh8+DC2bNlSwmkZJ+lqqDyam237REREVHrUajWaN3/y732jV3Dq1KmD+fPno3bt2tDr9Zg8eTIuXLjwr5IkIiIiKg1GB9OsWLECYWFh6N+/P6KiorBkyZLSzIuIiIioyIwWOE5OTvjpp59w+vRpLFiwAL6+vqWYFhEREVHRGb1FZWdnh6ZNm0KlUgEA7O3tDZY1Gk3pZEhERET0hIwOMt6/f7/RlUQEQUFBJZVTgTjImIiI6NlR7IOMX3rppX+VEBEREZG58Il9REREZHFY4BAREZHFYYFDREREFqdQBc6QIUPw6aefAgC8vb2LNNiHiIiIqLSYLHC+//57tG7dGq+99hoAIDMzE0uXLi3xxIiIiIiKyuRkmy1btkRgYCCio6MBANevX4etrW2JJ0ZERERUVCav4Dx48ABWVlYQyXlcjpubG7Kzs0s8MSIiIqKiMlngfPvtt9i8eTPc3d0xe/ZsHDlyBF9++WVp5EZERERUJCZvUa1btw5RUVEICgqCSqVCnz59kJiYWBq5ERERERWJyQLH29sbt2/fxrZt2wxeS01NLdHEiIiIiIrK5C2q7du3IywsDNu3b8e+ffvw559/YufOnYXqvEuXLkhMTERSUhKmTJlitF2/fv0gIggMDCx85kRERERGmLyC06RJE4PlgIAAjBkzxmTHVlZWWLp0KTp16oTz589DrVYjNDQUCQkJBu0qVKiACRMmICIi4glTJyIiIsrfEz/JWKPRoGXLlibbtWjRAsnJyUhJScGDBw8QEhKC3r1752k3a9YszJkzB3fv3n3SVIiIiIjyZfIKzsSJE5W/rays0KxZM1y4cMFkx56engbjdM6fP5+nMAoICIC3tzd27NiBDz/80GhfI0aMwMiRI01uk4iIiAgoRIHj5OSk/J2VlYXt27dj06ZN/3rDKpUKCxcuxLBhw0y2DQ4ORnBwMABA0tX/ettERERk2QoscKysrODk5FTg1RVj0tLS4O3trSx7eXkhLS1NWXZyckKjRo1w4MABAEC1atUQGhqKXr16ISoq6om3R0RERJSb5BfW1tYCQI4dO5bv+6bC2tpa/vjjD/H19RUbGxuJiYmRBg0aGG0fHh4ugYGBJvuVdHWR8mEwGAwGg/H0hVpdtO99o1dwIiMjERgYiJiYGGzduhW//vor/v77b+X9zZs3G1sVAPDw4UO899572LVrF6ytrbFixQrEx8dj5syZOHnypMFzdYiIiIiKkwo5lU4eUVFRCAwMxIoVK5TXRAQqlQoiguHDh5dWjgYkXQ2VR3OzbJuIiIhKl1qtRvPmT/69b/QKjru7OyZOnIjY2FilsHnk0cSbRERERGWR0QLH2toaFSpUMChsHmGBQ0RERGWZ0QInPT0ds2bNKs1ciIiIiIqF0ScZ53flhoiIiOhpYLTACQoKKs08iIiIiIqN0QInIyOjNPMgIiIiKjZPPNkmERERUVnHAoeIiIgsDgscIiIisjgscIiIiMjisMAhIiIii8MCh4iIiCwOCxwiIiKyOCxwiIiIyOKwwCEiIiKLwwKHiIiILA4LHCIiIrI4LHCIiIjI4rDAISIiIovDAoeIiIgsDgscIiIisjgscIiIiMjisMAhIiIii8MCh4iIiCwOCxwiIiKyOCxwiIiIyOKwwCEiIiKLwwKHiIiILA4LHCIiIrI4LHCIiIjI4rDAISIiIovDAoeIiIgsDgscIiIisjgscIiIiMjisMAhIiIii8MCh4iIiCwOCxwiIiKyOCxwiIiIyOKwwCEiIiKLwwKHiIiILA4LHCIiIrI4LHCIiIjI4rDAISIiIovDAoeIiIgsDgscIiIisjgscIiIiMjilGiB06VLFyQmJiIpKQlTpkzJ8/7EiRMRFxcHrVaLvXv3okaNGiWZDhERET1DpCTCyspKkpOTpWbNmmJjYyMxMTHi7+9v0KZjx45ib28vAGT06NESEhJisl9JV5dIvgwGg8FgMMpeqNVF+94vsSs4LVq0QHJyMlJSUvDgwQOEhISgd+/eBm0OHDiAO3fuAAAiIiLg5eVVUukQERHRM6TEChxPT0+kpqYqy+fPn4enp6fR9sOHD8fOnTvzfW/EiBFQq9VQq9XFnicRERFZnnLmTgAAhgwZgueffx4dOnTI9/3g4GAEBwcDACSdRQ4REREVrMQKnLS0NHh7eyvLXl5eSEtLy9MuKCgIH3/8MTp06ID79++XVDpERET0DCmxAketVsPPzw++vr5IS0vD4MGD8frrrxu0adq0KX788Ud07doVV65cKalUiOgZVKlSJbz//vvw9fWFSqUydzpEVAARwZkzZ/DNN98gIyOj+PotqejWrZucOnVKkpOTZdq0aQJAZs6cKa+88ooAkD179sjFixdFo9GIRqORrVu3muyTv6JiMBiFiUf/rbG2tjZ7LgwGo+CwtraWXr16ycyZM/O8V9RfUcHcO/WkwQKHwWAUJlavXs3ihsF4isLa2lpWr16d5/Uy9zNxIiJzUqlUePjwobnTIKJCevjwYbHeTmaBQ0RERBaHBQ4REZUpPj4+uH37NjQaDeLi4rB69WqUK/fPb2Latm2LEydOICEhAQkJCRgxYoTB+m+++Sb0ej10Oh2io6MxadKk0t4Fk3r37o1PP/3U3GkYNWDAAMTGxuLhw4cIDAw02s7YlEy+vr6IiIhAUlISQkJCYGNjAwAYO3Ys3n777RLP/xGz33d7kuAYHAaDUZj4+eefH3ttxmNhbN3Ax9q9YvZ9MRZWVlZm3b5KpSqRfn18fESv1yv7uG/fPnn99dcFgFStWlXOnj0rAQEBAkBcXV3l5MmT0r17dwEgXbt2laioKPHw8BAAYmtrK//5z3+KNb/iGNt19OhRcXV1LdVtPknUr19f6tatK+Hh4RIYGJhvm4KmZFq/fr0MGjRIAMgPP/wgo0ePFgBib28v0dHRRreb998tx+AQEZUpPj4+SEhIwMqVK3Hq1Cn873//Q1BQEI4cOYLTp0+jefPmAIDmzZvj2LFjiI6OxtGjR1G3bl0AgJWVFebNmwe9Xg+tVov33nsPAJCSkoKvv/4aUVFRePXVVzF48GDodDro9Xp8/fXX+ebi6OiIvXv3IioqCjqdDr169QIAfPXVVxgzZozSbvr06crVjsmTJyMyMhJarRYzZsxQ9ikxMRGrV69GbGwsvL298f3330OtViM2NlZpBwDdunVDQkICTp48icWLF2Pbtm0AAAcHByxfvhwnTpxAdHS0kosx2dnZiIyMVJ6EP3bsWKxatQoajQYAcO3aNfz3v//F1KlTAQAfffQRJk+ejPT0dADA/fv38dNPP+Xp193dHb/99htiYmIQExOD1q1bw8fHB3q9XmkzadIkTJ8+HQAQHh6ORYsWQa1W4+OPP8aZM2eU8SIODg44d+4cypUrh1q1amHnzp04efIkDh06hHr16uXZtp+fH+7du4dr164BAHr27ImIiAhER0djz549cHd3V87Hzz//jCNHjmDNmjVwc3PDxo0bERkZicjISLRp0waA8c/Qv5GYmIjTp08X2KagKZleeuklbNy4EQCwevVq9OnTBwBw584dnDlzRvn8l7RSrQr/bfAKDoPBKEyY+wqOj4+PPHjwQBo1aiQqlUpOnjwpy5cvFwDSq1cv2bx5swAQJycn5f/Og4KCZOPGjQLkTED866+/Ku9VqlRJAEhKSop8+OGHAkA8PDzk7Nmz4ubmJtbW1rJv3z7p3bt3nlysra3FyclJgJwrHklJSQJAmjZtKgcOHFDaxcXFiZeXl3Tq1El+/PFHAXKu0mzbtk1eeOEF8fHxkYcPH0rLli2VdR7lZWVlJeHh4dK4cWMpX768nDt3Tnx9fQWArFu3TrZt2yYA5IsvvpAhQ4YIAHFxcZFTp06Jg4NDnmP36ApO+fLlZf/+/dK4cWMBIJs2bZJevXoZtHd2dpZr164JALl27Zo4OzubPD8hISEyYcIEJXdnZ2eD7QKQSZMmyfTp0wWAhIeHy9KlS5X3tmzZIh07dhQAMnDgQAkODhYAsnfvXqlTp44AkBYtWsi+ffvybHvYsGEyf/58ZblixYrK38OHD1femz59upw8eVLs7OwEgKxdu1batm0rAMTb21vi4+ML/AzljgoVKiiPZHk8Hp8IO3cUdAWnf//+yn4DkDfeeEOWLFli8BkDIF5eXgbHddq0afLBBx8U8t9t0a/glImpGoiILFFKSgpiY2MBAHFxcdi3bx8AQK/Xw9fXFwDg4uKC1atXw8/PDyKijFV4+eWXsWzZMuWXYLkffrZ+/XoAOf/nfuDAAVy9ehUAsHbtWrRv3x5bt241yEOlUuHLL79E+/btkZ2dDU9PT1StWhUxMTFwd3eHh4cHqlSpgoyMDJw/fx4TJkxA586dlaskFSpUgJ+fH86dO4ezZ8/ixIkTSt8DBw7EyJEjUa5cOXh4eKBBgwawsrLCn3/+iTNnzgAAfvnlF4wcORIA0LlzZ/Tq1QuTJ08GANjZ2aFGjRpITEw0yLl27drQaDSoWbMmtm/fbnBlpTi89NJLeOuttwDkXCW6efMmKlWqVOA6j477o78HDRqEAwcOYPDgwfj+++/h6OiINm3a4Ndff1XalS9fPk8/Hh4eBg+39fLywvr16+Hh4QFbW1ukpKQo74WGhuLu3bsAcj4TDRo0UN5zdnaGo6Oj0c9Qbrdu3UJAQICpw1IqLl++jPr165f4dljgENEzYkYh20X9//j37t27p/ydnZ2tLGdnZyuDZmfNmoXw8HD069cPPj4+OHDggMl+//777wLfb9GiBX788UcAwGeffYbKlSujSpUqCAwMRFZWFlJSUmBnZwcA+PXXXzFgwABUq1ZN+QJXqVT46quv8H//938G/fr4+Bhs29fXF5MnT0bz5s1x/fp1rFy5UunXGJVKhf79+5u8/fHHH38gICAArq6uOHr0KF555RVs27YN8fHxCAwMRGhoqNI2MDAQcXFxAHIKycDAQISHhxfYf36ysrJgZfXPyI3H9yX3voeGhuLLL79EpUqVEBgYiP3798PR0RHXr183WUjcuXMHLi4uyvKSJUuwcOFCbNu2DR06dDC41Zd7m1ZWVmjVqpXB5woAvvvuO5OfoQoVKuDw4cP55vP6668jISGhwJzzY2xKpmvXrqFixYqwtrbGw4cP80zVZGdnhzt37jzx9p4Ux+AQEZmRi4uL8h//YcOGKa/v2bMHo0aNgrW1NQDke3UhMjISHTp0gKurK6ysrPDaa6/h4MGDiIyMREBAAAICArBt2za4uLjg8uXLyMrKQseOHZWrR0DOlYjBgwdjwIABypWHXbt24Z133oGjoyMAoHr16qhSpUqe7Ts7O+Pvv//GjRs34O7ujm7dugEATp06hVq1asHHxwcAMGjQIGWdXbt2Ydy4ccpy06ZNCzw+165dw9SpU/HRRx8BAJYuXYphw4bhueeeAwBUrlwZc+bMwdy5cwHkjCuaN28eqlatCgCwsbHB8OHD8/S7b98+vPvuuwByCgdnZ2dcunQJ7u7uqFy5MmxtbdGzZ0+jef39999Qq9VYvHgxwsLCkJ2djczMTKSkpGDAgAFKuyZNmuRZNyEhAXXq1FGWc38Ghg4danSbu3fvNjh2j46Bsc9Qbo+u4OQXRSluAMMpmWxsbDB48GCl8AwPD1eOw9ChQw2uKtatW1e5slmSWOAQEZnR3Llz8dVXXyE6Otrgp9A//fQTzp07B51Oh5iYmDxz+QHAxYsXMXXqVISHh0Or1SIqKsrgysYja9euxfPPPw+dToe33nrL4AstPj4eTk5OSEtLw8WLFwHkFFfr1q3D8ePHodPpsHHjRjg5OeXpV6fTQaPRIDExEevWrcPRo0cBAHfv3sWYMWPw+++/4+TJk8jMzMSNGzcA5FyxsrGxgU6nQ2xsLGbNmmXyGG3ZsgUODg5o164dLl68iDfeeAPBwcFISEjAsWPHsGLFCoSFhQEAdu7cie+++w579+5FbGwsoqOj4ezsnKfPCRMm4MUXX4ROp0NUVBQaNGiArKwsfP7554iMjMSePXvy3DZ73Pr16/Hmm28a3LoaMmQIhg8fjpiYGMTFxSmDbnM7dOiQwVWeGTNm4Ndff8XJkyeV2435GT9+PJ5//nlotVrExcVh9OjRAIx/hv6NPn36IDU1Fa1bt8b27dvx+++/A8i5vbZ9+3YAOQ/me++997Br1y4kJCRgw4YNiI+PBwBMmTIFH3zwAZKSkuDq6orly5crfbdt2xZ79uwpljxNKdLgHXMFBxkzGIzCRH6DFRmlF46OjsrfS5culffff9/sOZWl+OabbyQoKMjseZR2NG3atMB/m/yZOBERlWkjRoxQHtTn4uKijAmiHF9++SUcHBzMnUapc3NzK7UHHHKQMRERFbtvvvkG33zzjbnTKLMuX76sPBvoWbJ3795S2xav4BAREZHFYYFDREREFocFDhEREVkcFjhERERkcVjgEBGVkKysLGg0Guj1eoSGhho8vbZBgwbYt2+fMqnhJ598YrBu165doVarERcXh+joaMyfP7+00y+ydevWQavV4v333y9U+8zMzBLLZfHixUhKSoJWqzX6hGE7OzscOHDA4CnGZUnlypWxf/9+ZGZmYsmSJUbbVapUCbt378bp06exe/duVKxYUXkvv+Pg5uaGnTt3lnT6ZmX238U/SfA5OAwGozBRFp6Dk5mZqfy9atUqmTZtmgAQOzs7SU5Olk6dOgkAsbe3lx07dsiYMWMEgDRs2FCSk5OlXr16AuRMBjl69Ohize3R5IzFHVWrVjWYaPFJj1NxRrdu3WTHjh0CQFq2bCkRERH5thszZoyMHz/+ifpWqVSl9jlycHCQtm3byqhRo2TJkiVG282ZM0emTJkiAGTKlCny9ddfmzwOK1askDZt2pTavpiK4nwODsy9M08aLHAYDEZhwuA/lCIlEyZyyP3FPWrUKGU26nfeeUdWr15t0LZWrVpy7tw5ASCrV6+Wt99+22T/jo6OsmLFCtHpdKLVaqVfv355ttu/f39ZuXKlAJCVK1fKDz/8IBEREbJgwQJJSUkRFxcXpe3p06fF3d1d3NzcZOPGjRIZGSmRkZH5fgGWL19e2XZ0dLQys7ZWq5Xbt2+LRqORdu3aGazj7u4uv/32m8TExEhMTIy0bt3aIF9HR0fZu3evREVFiU6nU2YNd3BwkLCwMImJiRG9Xi8DBw4UAPLVV19JXFycaLVamTdvXp4cly1bJoMHD1aWExMTpVq1annaHT16VHx8fArMwcfHRxITE2X16tUSGxsrNWrUkMmTJ0tkZKRotVqZMWOG0t/mzZvl5MmTEhsbKyNGjCi2z/TQoUMLLHBy71+1atUkMTHR5HHo1auXwSzp5g7OJk5E9BSxsrJCUFCQ8rj6hg0bIirKcELPP//8ExUqVICTkxMaNWqEBQsWmOz3008/xY0bN5T5jnLfkjDGy8sLbdq0QXZ2NqytrdG3b1+sWrUKLVq0wNmzZ3H58mWsXbsWixYtwtGjR+Ht7Y1du3YZzGINAGPHjoWIoEmTJqhXrx52796NunXrolevXggLC8v3dtC3336LgwcPol+/frCyskKFChUM3r979y769u2LzMxMuLq6IiIiAqGhoejatSsuXLigzA3l7OyMypUro2/fvsqs1Llv/z3i6emJ1NRUZfn8+fPw9PRUpqQAcuaqqlWrFs6ePVtgDgDg5+eHoUOH4sSJE+jUqRP8/PzQokULqFQqhIaG4oUXXsDhw4fxzjvvICMjA3Z2dlCr1di0aRP++usvg9wWLlyIF198MU/OISEhmDNnjvETWICqVasq+3bx4kVlPq6CjsPJkycxe/bsIm2vrGOBQ0SWT6Uyy2bt7e2h0Wjg6emJhISEYp9/5+WXX8bgwYOV5evXr5tc59dff0V2djaAnLmUPvvsM6xatQqDBw9W5lR6+eWXDQoaZ2dnODo6Gsxs3a5dO2U8yKlTp3D27FnUrVsXN2/eNLrtl156CW+99RaAnBnVH2+rUqnw5Zdfon379sjOzoanpyeqVq0KvV6PBQsW4Ouvv0ZYWBiOHDkCa2tr3L17F8uXL0dYWJgyF9WTcnNzMzhuxnIAgLNnz+LEiRMAgM6dO6Nz587QaDQAcmbr9vPzw+HDhzF+/Hj07dsXAODt7Q0/Pz9lvUc++OCDIuX7JHIuMhbs8uXLqF69eonnYg5lc0QVEZEFuHPnDgICAuDj4wOVSoWxY8cCyJngMjAw0KBtzZo1cevWLWRmZiIuLi7P+08i9xebnZ2dwXu5i5Tjx4+jTp06cHNzQ58+ffDbb78ByLni1KpVK2W2aS8vL4P1SsqQIUNQpUoVBAYGIiAgAJcuXYKdnR2SkpLQrFkz6PV6zJ49G59++ikePnyIFi1aYOPGjejZs6cyGWRuaWlp8Pb2Vpa9vLyUWbcfuXPnjsExMpYDYHjsVCoVvvrqK+UY+fn5YcWKFejQoQNefvlltG7dGk2bNoVGo8lzDoCcKzgajSZPTJkypcjH79KlS6hWrRoAoFq1arh8+bLJ42BnZ4c7d+4UeZtlGQscIqISdufOHYwfPx6TJk2CtbU11q5di3bt2iEoKAhAzpfMt99+i7lz5wIA5s2bh2nTpsHPzw9AzpfpqFGj8vS7Z88epWgC/rlFdenSJdSvXx8qlUq5kmDM5s2bsXDhQiQkJCi3UXbv3o1x48YpbZ577rk86x0+fBhDhgwBkHPrpkaNGjh16lSB29q3bx/effddADlF1OOzfLu4uODy5cvIyspCx44d4evrCyBnBuvbt29j7dq1mDdvHpo1awZHR0e4uLhg586dmDhxYr45hoaGKleMWrZsiRs3bhjcngJyrnpZW1ujfPnyBebwuF27duGdd96Bo6MjAKB69eqoUqUKXFxckJGRgTt37qBevXpo1apVvut/8MEHSnGUO4p6e+rR/g4dOhQAMHToUGzdutXkcahbty5iY2OLvM2yzuyDip4kOMiYwWAUJsrar6gASGhoqLzxxhsCQBo1aiTh4eGSmJgoSUlJ8tlnnxm07dGjh5w8eVLi4+MlLi5O5syZk6d/R0dHWbVqlej1eomJiZG+ffsKkDOwODk5WY4fPy5LliwxGGTcv39/gz4CAwNFROStt95SXnN1dZWQkBDRarUSFxcnP/zwQ55tGxtk7OPjI3q9Pt/j4e7uLlu2bBGdTicajUZatWplcJxcXV3l2LFjotPpZMWKFRIfHy8+Pj7SuXNn0Wq1otFoJDIyUgIDA6VatWpy4sQJ0Wq1otPpDPLPHd99950kJyeLTqeTwMDAfNv89NNPyszexnLIb7/Gjx8vOp1OdDqdHDt2TGrVqiW2trayY8cOiY+Pl82bN0t4eLh06NDhX3+WUlJS5Nq1a5KZmSmpqani7+8vACQ4OFjZr8qVK8vevXvl9OnTsmfPHqlUqZLJ4zBp0iR57733zP5v5VHwV1RlIA8Gg1G2oywUOIynIwICAp7Zz8vBgwelYsWKZs/jURRngcNbVERE9EzTaDQIDw8vsw/6Kylubm5YuHBhoQanP434KyoiInrmrVy50twplLqrV68q43Qs0bNVrhIREdEzgQUOERERWRwWOERERGRxWOAQERGRxWGBQ0RUQrKysqDRaKDX6xEaGmowX1KDBg2wb98+JCYm4vTp0/jkk08M1u3atSvUajXi4uIQHR2N+fPnl3b6RbZu3TpotVq8//77hWqfmZlZInnUq1cPx44dw927dzFp0qQC2+7btw9OTk4lkkdx2LlzJzIyMrBt2zajbWxtbRESEoKkpCRERETAx8dHeW/q1KlISkpCYmIiOnfuDCBnHq6DBw/C2tq6xPM3F7P/7v1Jgs/BYTAYhYmy8FyT3A/6W7VqlUybNk0AiJ2dnSQnJ0unTp0EgNjb28uOHTtkzJgxAkAaNmwoycnJUq9ePQEgVlZWMnr06GLNzdraukT2uWrVqpKUlFTk41ScUaVKFXn++edl9uzZMmnSJKPtunfvLgsXLnyivq2srEr1s/TSSy9Jz549Zdu2bUbbvPvuu8pDGQcNGiQhISECQPz9/SUmJkZsbW3F19dXkpOTlfw/++wzef3110t1XwoKPgeHiOgJyPySiSdx/PhxeHp6AgBef/11HD16VJl8886dO3jvvfcwdepUAMB///tffPHFF8rUB9nZ2Vi2bFmePh0dHbFixQrodDpotVr069cPgOEVkf79+ys/gV65ciV++OEHREREYO7cuUhJSTG4qnT69Gm4u7vDzc0NGzduRGRkJCIjI9GmTZs82y5fvryy7ejoaHTs2BFAzjQPnp6e0Gg0aNeuncE67u7u+O233xATE4OYmBi0bt06z/7s3bsXUVFR0Ol06NWrFwDAwcEBYWFhiImJgV6vx8CBAwEAX331FeLi4qDVajFv3rw8OV65cgUnT57EgwcP8j0njwwZMsTg59KbN2/GyZMnERsbixEjRiivZ2ZmYv78+UruQ4YMwYkTJ6DRaLBs2TLlOTrff/891Go1YmNjMWPGjAK3XVj79+83eaWrd+/eWL16NQBg48aNylQgvXv3RkhICO7fv48zZ84gOTkZLVq0AABs2bJFmXLD0vA5OEREJczKygpBQUFYvnw5AKBhw4aIiooyaPPnn3+iQoUKcHJyQqNGjbBgwQKT/X766ae4ceMGmjRpAuCfuagK4uXlhTZt2iA7OxvW1tbo27cvVq1ahRYtWuDs2bO4fPky1q5di0WLFuHo0aPw9vbGrl27DGYXB4CxY8dCRNCkSRPUq1cPu3fvRt26ddGrVy+EhYUhICAgz7a//fZbHDx4EP369YOVlRUqVKhg8P7du3fRt29fZGZmwtXVFREREQgNDUXXrl1x4cIF9OzZE0DO7OaVK1dG3759Ub9+fQAwKNSeVNu2bQ3m+nrnnXeQkZEBOzs7qNVqbNq0CX/99RcqVKiAEydOYPLkyahfvz6mTJmCtm3bIisrC0uXLsWQIUOwZs0afPzxx8jIyICVlRX27duHxo0bQ6/XG2xz8uTJ+RYWhw4dwoQJE4q0H56enkhNTQUAPHz4EDdu3ICrqys8PT0RERGhtDt//rxSbMfGxqJ58+ZF2l5ZxwKHiCyearJ5tmtvbw+NRgNPT08kJCQoV2yKy8svv4zBgwcry4V5Iu2vv/6K7OxsAMD69evx2WefYdWqVRg8eDDWr1+v9Ju7oHF2doajo6PBbNrt2rXDkiVLAACnTp3C2bNnUbduXdy8edPotl966SVl0sfs7Ow8bVUqFb788ku0b98e2dnZ8PT0RNWqVaHX67FgwQJ8/fXXCAsLw5EjR2BtbY27d+9i+fLlCAsLQ1hYmMl9N6Zy5cq4deuWsjx+/HhlklJvb2/4+fnhxIkTyMrKwqZNmwAAQUFBCAwMhFqtBpBzrh/N3j1w4ECMHDkS5cqVg4eHBxo0aJCnwJk/f36ZGFeVnZ2N+/fvo0KFCgbHwBLwFhURUQm5c+cOAgIC4OPjA5VKpcz8HR8fj8DAQIO2NWvWxK1bt5CZmYm4uLg87z8JEVH+trOzM3gvd5Fy/Phx1KlTB25ubujTpw9+++03ADlXnFq1aqXMcO3l5WWwXkkZMmQIqlSpgsDAQAQEBODSpUuws7NDUlISmjVrBr1ej9mzZ+PTTz/Fw4cP0aJFC2zcuBE9e/bE77//XuTtZmVlQaVSAQA6dOiAl19+Ga1bt0bTpk2h0WiUY3j37l2lOFSpVFi9erVyjOrXr4+ZM2fC19cXkydPRlBQEJ577jls3749zzkAcq7gaDSaPLF48eIi70daWhq8vb0BANbW1nBxccG1a9cMXgdyruKlpaUpy+XLl8fdu3eLvN2yigUOEVEJu3PnDsaPH49JkybB2toaa9euRbt27ZQxEnZ2dvj2228xd+5cAMC8efMwbdo0+Pn5Acj5Ms19C+WRPXv2KEUT8M8tqkuXLqF+/fpQqVTKlQhjNm/ejIULFyIhIQF//fUXgJxxNOPGjVPaPPfcc3nWO3z4sHKLxc/PDzVq1FDGDBmzb98+vPvuuwByiihnZ2eD911cXHD58mVkZWWhY8eO8PX1BQB4eHjg9u3bWLt2LebNm4dmzZrB0dERLi4u2LlzJyZOnJhvjoV16tQp1KpVS8khIyMDd+7cQb169dCqVSuj+zJgwABUqVIFAFCpUiXUqFEDzs7O+Pvvv3Hjxg24u7ujW7du+a4/f/58pTjKHUW9PQUAoaGhGDp0KABgwIAB2L9/v/L64MGDYWtrC19fX/j5+SEyMhJAztWrq1evIisrq8jbLcvMPmr6SYK/omIwGIWJsvYrKgASGhoqb7zxhgCQRo0aSXh4uCQmJkpSUpJ89tlnBm179OghJ0+elPj4eImLi5M5c+bk6d/R0VFWrVoler1eYmJipG/fvgJA+vfvL8nJyXL8+HFZsmSJrFy5UgDIypUrpX///gZ9BAYGiojIW2+9pbzm6uoqISEhotVqJS4uTvllTu4oX768rFixQnQ6nURHR0vHjh0FgPj4+Iher8/3eLi7u8uWLVtEp9OJRqORVq1aGRwnV1dXOXbsmOh0OlmxYoXEx8eLj4+PdO7cWbRarWg0GomMjJTAwECpVq2anDhxQrRareh0OoP8H0XVqlUlNTVVbty4IRkZGZKamipOTk552n3yyScyfPhwASC2trayY8cOiY+Pl82bN0t4eLh06NAh3/M5cOBA0Wg0otVq5eTJk9KyZUvlOJ86dUr27t0rmzZtkqFDh/7rz9KhQ4fk8uXLcvv2bUlNTZXOnTsLAJk5c6a88soryjnZsGGDJCUlyYkTJ6RmzZrK+tOmTZPk5GRJTEyUrl27Kq/3799f5s+fb/Z/K4+iOH9FBXPvzJMGCxwGg1GYKAsFDuPpiGrVqsnu3bvNnoc5YtOmTeLn52f2PB4FfyZORERUTC5evIjg4OAy/aC/kmBjY4MtW7YgKSnJ3KmUCP6KioiInnm//vqruVModQ8ePMCaNWvMnUaJ4RUcIrJIImLRj6AnsjTW1tYGvwD8t1jgEJFFOnPmDHr06MEih+gpYG1tjR49euDMmTPF1idvURGRRfrmm2/w/vvvo3///sozToiobBIRnDlzBt98802x9ckCh4gsUkZGBqZPn27uNIjITEr0FlWXLl2QmJiIpKQkTJkyJc/7BU3tTkRERFRUJVbgWFlZYenSpejWrRsaNGiA1157Df7+/gZthg8fjoyMDPj5+WHRokWYM2eOyX7PXS+hhImIiMhilFiB06JFCyQnJyMlJQUPHjxASEgIevfubdDG2NTuBbliWXOBERERUQkosTE4uadtB3KmZ2/ZsqXRNrmndr927ZpBuxEjRmDkyJEAgIaNGimzt1LZ4ebmhqtXr5o7DXoMz0vZxPNSNvG8lE316tUr0npPxSDj4OBgBAcHAwDUajWaN29u5ozocTwvZRPPS9nE81I28byUTUW9qFFit6hMTc/+eJvcU7sTERER/RslVuCo1Wr4+fnB19cXNjY2GDx4MEJDQw3aGJvanYiIiOjfKrFZQbt16yanTp2S5ORkmTZtmgCFn9rdWIwYMcLss50yeF6eluB5KZvB81I2g+elbEZRz4vq//9BREREZDE4FxURERFZHBY4REREZHHKbIHDaR7KJlPnZeLEiYiLi4NWq8XevXtRo0YNM2T57DF1Xh7p168fRASBgYGlmN2zqzDn5dVXX0VcXBxiY2Oxdu3aUs7w2WPqnHh7e2P//v2Ijo6GVqtFt27dzJDls2f58uW4dOkS9Hq90TaLFy9GUlIStFotAgICCtWv2QcQPR5WVlaSnJwsNWvWFBsbG4mJiRF/f3+DNu+++6788MMPAkAGDRokISEhZs/b0qMw56Vjx45ib28vAGT06NE8L2XkvACQChUqyMGDB+X48eMSGBho9rwtPQpzXurUqSPR0dFSsWJFASBVqlQxe96WHIU5Jz/++KOMHj1aAIi/v7+kpKSYPe9nIV544QUJCAgQvV6f7/vdunWTHTt2CABp2bKlREREmD7fKINKapoH+ncKc14OHDiAO3fuAAAiIiLg5eVljlSfKYU5LwAwa9YszJkzB3fv3jVDls+ewpyXESNGYOnSpbh+/ToA4MqVK2bI9NlRmHMiInB2dgYAuLi44MKFC+ZI9Zlz+PBh/PXXX0bf7927N37++WcAwIkTJ1CxYkVUq1atwD7LZIGT3zQPnp6eRtvknuaBSk5hzktuw4cPx86dO0sjtWdaYc5LQEAAvL29sWPHjtJO75lVmPNSt25d1K1bF0eOHMHx48fRpUuX0k7zmVKYczJjxgy88cYbSE1NxY4dOzBu3LjSTpPy8aTfP8BTMlUDPX2GDBmC559/Hh06dDB3Ks88lUqFhQsXYtiwYeZOhR5Trlw5+Pn5oWPHjvDy8sKhQ4fQuHFj3Lhxw9ypPbNee+01rFq1CgsXLkSrVq2wZs0aNGrUCCJi7tToCZXJKzic5qFsKsx5AYCgoCB8/PHH6NWrF+7fv1+aKT6TTJ0XJycnNGrUCAcOHEBKSgpatWqF0NBQDjQuYYX593L+/HmEhoYiKysLZ86cwenTp+Hn51faqT4zCnNOhg8fjg0bNgDIuc1uZ2cHNze3Us2T8irs98/jzD646PGwtraWP/74Q3x9fZWBYA0aNDBoM2bMGINBxuvXrzd73pYehTkvTZs2leTkZKlTp47Z831WojDnJXeEh4dzkHEZOS9dunSRVatWCQBxdXWVc+fOSeXKlc2eu6VGYc7Jjh07ZOjQoQJA6tevL2lpaWbP+1kJHx8fo4OMu3fvbjDI+MSJE4Xp0/w7lV+UxDQPjJI/L3v27JGLFy+KRqMRjUYjW7duNXvOz0KYOi+5gwVO2TovCxYskLi4ONHpdDJo0CCz52zpYeqc+Pv7y5EjRyQmJkY0Go106tTJ7Dk/C7Fu3Tq5cOGC3L9/X1JTU+Wdd96RUaNGyahRo5Q23333nSQnJ4tOpyvUf8M4VQMRERFZnDI5BoeIiIjo32CBQ0RERBaHBQ4RERFZHBY4REREZHFY4BAREZHFYYFD9IzJysqCRqNRwsfHx2jbzMzMf729lStX4s8//4RGo0FUVBRatWr1xH0EBwfD398fAPDRRx8ZvHf06NF/nSPwz3HR6/UIDQ2Fi4tLge2fe+45zjRNVMaZ/ffvDAaj9CIzM7NE2hqLlStXSv/+/QWAdOrUSbRabanlX9R+V61apTwjxVgMHTpUlixZYvbzyWAw8g9ewSF6xjk6OmLv3r2IioqCTqdDr1698rSpVq0aDh48qFzhaNeuHQCgU6dOOHbsGKKiorBhwwY4OjoWuK1Dhw6hTp06AICJEydCr9dDr9djwoQJAAAHBweEhYUhJiYGer0eAwcOBACEh4cjMDAQX331Fezt7aHRaPC///0PwD9XmX755Rd0795d2dbKlSvRv39/WFlZYe7cuYiMjIRWq8XIkSNNHpPjx48rE/k1b94cx44dQ3R0NI4ePYq6devCxsYGn3/+OQYNGgSNRoOBAwfCwcEBy5cvx4kTJxAdHZ3vcSSi0mX2KovBYJReZGVlKU+a/u2338Ta2lqcnJwEyJkuICkpSWn76KrGBx98oFzRsLKykgoVKoirq6scPHhQHBwcBID897//lU8//TTP9nJfwRkwYIBERERIs2bNRKfTiYODgzg6OkpsbKw0bdpU+vXrJ//3f/+nrOvs7CyA4dOXH7+C82i5T58+yrQHNjY2cu7cObGzs5MRI0bIxx9/LADE1tZW1Gq1+Pr65snzUT9WVlayYcMG6dKliwAQJycnsba2FgASFBQkGzduFCDvFZwvvvhChgwZIgDExcVFTp06pRwbBoNR+sHZxImeMXfu3EFAQICyXK5cOXz55Zdo3749srOz4enpiapVq+LSpUtKG7VajRUrVsDGxgZbtmyBVqtFhw4d0KBBA2UMjK2tLY4fP57vNufNm4dPPvkEV65cwfDhwxEUFITNmzfj9u3bAIDffvsNL7zwAn7//XcsWLAAX3/9NcLCwnDkyJFC79fOnTuxePFi2NraomvXrjh06BDu3r2Lzp07o0mTJhgwYAAAwMXFBX5+fjhz5ozB+o+uDHl6eiIhIQF79uxR2q9evRp+fn4QEdjY2OS7/c6dO6NXr16YPHkyAMDOzg41atRAYmJiofeBiIoPCxyiZ9yQIUNQpUoVBAYGIisrCykpKbCzszNoc/jwYbRv3x49evTAqlWrsHDhQmRkZGDPnj14/fXXTW7jww8/xKZNm5TloKCgfNslJSWhWbNm6N69O2bPno19+/Zh1qxZhdqPe/fu4cCBA+jSpQsGDRqEkJAQAIBKpcK4ceOwe/fuAtd/VPjZ29tj165dGDt2LJYsWYJZs2YhPDwc/fr1g4+PDw4cOJDv+iqVCv3798fp06cLlS8RlSyOwSF6xrm4uODy5cvIyspCx44d4evrm6dNjRo1cOnSJfz000/46aef0KxZM0RERKBt27aoXbs2gJzxM35+foXa5uHDh9GnTx/Y29vDwcEBffv2xeHDh+Hh4YHbt29j7dq1mDdvHpo1a5Zn3QcPHqBcufz/32z9+vV4++23latBALBr1y68++67yjp+fn5wcHAwmtudO3cwfvx4TJo0CdbW1nBxcUFaWhoAYNiwYUq7zMxMODk5Kcu7du3CuHHjlOWmTZsW6lgQUckx+30yBoNRevH4GBZXV1c5duyY6HQ6WbFihcTHx4uPj49B27feekv0er1ER0fLoUOHlDEsL774okRGRopWqxWtVpvv7OW5x+DkjokTJ4perxe9Xi8TJkwQANK5c2fRarWi0WgkMjJSGXeTewzO119/LfHx8fK///0vz/6UK1dOrl27JitWrFBeU6lU8sUXX4hOpxO9Xi/79+9XxvYUdFxCQ0PljTfekFatWsmpU6ckOjpaZs2aJSkpKQJAKlWqJJGRkaLRaGTgwIFiZ2cny5YtE51OJ7GxsbJt2zazn2sG41kOziZOREREFoe3qIiIiMjisMAhIiIii8MCh4iIiCwOCxwiIiKyOCxwiIiIyOKwwCEiIiKLwwKHiIiILM7/A+gjpOtGUjnAAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 576x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "mobilenet = load_model('ModelWeights/Mobilenet_Masks.h5')\n",
    "Viz.plot_roc_auc(mobilenet, x_val, y_val, 'dark_background', 'mask', path = 'Images/Mobilenet_ROC.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:plaidml:Opening device \"opencl_amd_ellesmere.0\"\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "mobilenet = load_model('ModelWeights/Mobilenet_Masks.h5')\n",
    "y_test_prob = [np.argmax(i) for i in mobilenet.predict(x_test)]\n",
    "y_test_labels = [np.argmax(\n",
    "    i) for i in y_test]\n",
    "test_cnf = confusion_matrix(y_test_labels, y_test_prob)\n",
    "\n",
    "val_prob = [np.argmax(i) for i in mobilenet.predict(x_val)]\n",
    "val_labels = [np.argmax(i) for i in y_val]\n",
    "val_cnf = confusion_matrix(val_labels, val_prob)\n",
    "\n",
    "#this function creates a confusion matrix given the confusion matrixes of test and train\n",
    "Viz.plot_model_cm(test_cnf, val_cnf, classes = ['No Mask', 'Mask'], theme = 'dark_background', model_type = 'mask', \n",
    "             path = 'Images/Mobilenet_CM.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "assert False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "%%javascript\n",
    "IPython.notebook.save_notebook()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "os.system(r'%windir%\\system32\\rundll32.exe powrprof.dll,SetSuspendState Hibernate')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "FaceMaskEmotionDetection",
   "language": "python",
   "name": "facemaskemotiondetection"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}