model.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "import keras\n",
    "import cv2\n",
    "from keras import layers\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "from tensorflow.keras.utils import to_categorical\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## data generation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# path of images for training\n",
    "train_path = \"./images/training_images\" \n",
    "# path of images for validation\n",
    "valid_path = \"./images/validation_images\" \n",
    "# path of images for testing\n",
    "test_path = \"./images/test_images\" "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 300 images belonging to 2 classes.\n",
      "Found 20 images belonging to 2 classes.\n",
      "Found 40 images belonging to 2 classes.\n"
     ]
    }
   ],
   "source": [
    "# ImageDataGenerator is used for generating data\n",
    "# flow_from_directory() generates batches of normalized tensor image data from the respective data directories.\n",
    "# we provide path for directory in flow_from_directory() function\n",
    "# target_size is size of the image and we use gray scale because many operations perform better in grayscale images\n",
    "# class_mode will devide images into two class (open,closed)\n",
    "# shuffle will shuffle the data everytime\n",
    "# batch_size is number of batches we want to make from directory images\n",
    "train_batches = ImageDataGenerator().flow_from_directory(\n",
    "    directory=train_path, \n",
    "    target_size=(24,24),\n",
    "    class_mode='categorical',\n",
    "    shuffle=True, \n",
    "    batch_size=10, \n",
    "    color_mode='grayscale')\n",
    "valid_batches = ImageDataGenerator().flow_from_directory(\n",
    "    directory=valid_path, \n",
    "    target_size=(24,24),\n",
    "    class_mode='categorical',\n",
    "    shuffle=True, \n",
    "    batch_size=2, \n",
    "    color_mode='grayscale')\n",
    "test_batches = ImageDataGenerator().flow_from_directory(\n",
    "    directory=test_path, \n",
    "    target_size=(24,24),\n",
    "    class_mode='categorical',\n",
    "    shuffle=True, \n",
    "    batch_size=10, \n",
    "    color_mode='grayscale')\n",
    "\n",
    "# We have 300 images in training_images , 20 images in validation_images and 40 images in test_images "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'closed': 0, 'open': 1}\n",
      "{'closed': 0, 'open': 1}\n",
      "{'closed': 0, 'open': 1}\n"
     ]
    }
   ],
   "source": [
    "# It has assigned 0 index to closed and 1 to open for every directory\n",
    "print(test_batches.class_indices) \n",
    "print(train_batches.class_indices)\n",
    "print(valid_batches.class_indices)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# collecting labels for test images\n",
    "test_label = []  \n",
    "count = 0\n",
    "for i,j in test_batches:\n",
    "    if count == 4:   # because we have total 40 images and 10 images per batch , so there will be 4 batch\n",
    "        break        # if we will not break this , it will go into an infinite loop\n",
    "    count+=1\n",
    "    for k in j:\n",
    "        test_label.append(k)  # appending labels to test_label"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## data visualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(40, 2)\n",
      "<class 'numpy.ndarray'>\n",
      "<class 'numpy.ndarray'>\n"
     ]
    }
   ],
   "source": [
    "test_label = np.array(test_label) # defining numpy array for labels because faster and easier execution\n",
    "print(test_label.shape)  # it contains array for all 40 images, [0,1] for open eye and [1,0] for closed eye \n",
    "print(type(test_label[0])) # test_label is a 2d numpy array\n",
    "print(type(test_label))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10, 24, 24, 1)\n",
      "[[1. 0.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [1. 0.]]\n"
     ]
    }
   ],
   "source": [
    "for x,y in train_batches: # train_batches contains images and corresponding labels for that images\n",
    "    print(x.shape) # x has 10 gray images (one batch) of size (24,24)\n",
    "    print(y) # y contains label for one batch\n",
    "    break # if we do not break , it will go into infinite loop"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0. 1.]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0. 1.]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0. 1.]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0. 1.]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1. 0.]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAaAAAAGfCAYAAAAZGgYhAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjEsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvc2/+5QAAAAlwSFlzAAAPYQAAD2EBqD+naQAANFBJREFUeJzt3X+MHdV5//EzM/fX7voXhsQuwm6REkFSBFGcBCj9kYIbK4oiKPyRSpHqpihRKCDASG0sNUStWpkGKRBSQ6KUgio1dUVVEhGptMgJRlVtCk5RCWmsVqLFqbFJ+q1/rff+mpmvZhxvvXjP59k918tZe98v6YZ4z87cuefOvc/Ovc9znqQsy9IBAPA2S9/uOwQAoEIAAgBEQQACAERBAAIAREEAAgBEQQACAERBAAIAREEAAgBEQQACAERBAAIARNFYqB1v377dPfDAA+7gwYPuqquucl/5ylfchz70IXO7oijcgQMH3PLly12SJAt1eACABVKt8Hbs2DF38cUXuzQV1znlAtixY0fZarXKP//zPy9fffXV8tOf/nS5atWq8tChQ+a2+/fvr9am48aNGzdu7ty+Ve/nSlL9z9mOfldffbX74Ac/6P70T/90+qpm3bp17s4773Sf+9zn5LZHjhxxq1atcr901T2ukbXPGC8amdw+H/df1A3H9LaDcX+kLsW1YpLL3bok909x2dBXeUW2MFeBSeE/psQ4I0pxSGXqH8w7er/Djn/b4ZjeNm+LY2qUwR9CF+J5L9p6okoxkUnhf6xlauy3U/j329EnY6M59I51Ov6xtcuPyv3+zJh/fKLRl9t2soF3rCleXGqssizrecfGUv9YZU3D/3jWZP6x3OnX66D0vwf1yqbcti+2HYg3qIHT73tKJ/E/N5XUzX4unjieu0//0g/d4cOH3cqVK9++j+D6/b7bu3ev27p16/8dZJq6jRs3ut27d5/x+71er76dUl221QeWtV0j68w7ACUN8ZCaetuy6X83KsS5kfpftyePSbyhFEYASozxBQlA/ve3WpmGBSDXMvbbFm/KIsDURHArRghA8hViBSDxvCe5eKzZwgWgtOU/WbOO/82mMaGfgNa4/8ltqfmvplG8LJtJGjRW6WT+uRhL9TyNi/eKiSxdkACUibFKQwagsDFLJ9VvBqnxeK2vUc56EsJPfvITl+e5W7NmzYyfV/+uvg96q23bttUR8tStulICAJz/omfBVVdK1cdup2779++PfUgAgLfBWf8I7qKLLnJZlrlDhw7N+Hn177Vr157x++12u74BAJaWsx6AWq2W27Bhg9u5c6e76aabppMQqn/fcccdc99R9dnhbJ8fJgvzvUclHarxJHA76zuVs54DcvI+rd2W4dvKh6O+eB/qJy/riWQNKyVfDKv7NT8el+eMPib1fZj6/rw0Ek9E/oKzUooG4rsnlV3y/5rjcr8t8X1LT2VyOOdWtya9Y0PxPU+RGudT0gr+rmY89ycpLE+n/PdpvJ5T8WbQcfoL/2bi//5uUPrnv2skN/gSCSoTRrKGLxGkmRlfJC9kHdCWLVvc5s2b3Qc+8IG69uehhx5yk5OT7lOf+tRC3B0A4By0IAHoE5/4hPvxj3/s7rvvvjrx4H3ve5975plnzkhMAAAsXQu2EkL1cdu8PnIDACwp0bPgAABLEwEIABAFAQgAEAUBCABwfiUhjKqqoZi1HsKqB1H1LUa9TmOqCKszMRYjVcUZ5mKkTVW3McIxyQVFjU3FMan6ioZRh+W6/qFMjNX3OxU2h0aJit7WWtsuDasNU4u9nrxff/FS0dJPnnq8w7Z/8MddPVGTXf9kvGP5pD6mCf8DHhMLlbYzvQjjcbFCbWoUuxXiycvEthdmx+V+V6f+8ZaxCKM65oE43smyFbzg6IVGHZBv+cZjjbnVAXEFBACIggAEAIiCAAQAiIIABACIggAEAIiCAAQAiGJxp2Eby63Pt+VC1h+h9YFMeQ7fbyHagFutBNT9Wi0V1JL/IfM+va1KVzcyM1PxeLKBPqa8L9KlxWr0Zhp2IzBFvqKG1Vwk4cdkp/X7x3KR3j2c0ufpia5/24N5+N+5nYY/RXi80Q9OW7bTsEX7FXEiZ7pZiWzl0DJeILLVwwhtHpaLNOyLMt2vZNzT8mJsju0YuAICAERBAAIAREEAAgBEQQACAERBAAIAREEAAgBEQQACAESxaOuAQlobVNK+6ENg1MYkat9FeM2NvM9C59mneWBNTjFCHZAYq8eTt//vFqu9RCaeu1Rsa0y/rskxzkW52zK8HUMpjrnIRqkD8m87PKH3O5zyv5X08nG57X8P/Q+o2fK3XGg3dTuGZsP/xHcaw+A6oIUybrQ+aIkXQTMZhtUPOecmRFuLzChKayazP3dNq23OT3EFBACIggAEAIiCAAQAiIIABACIggAEAIiCAAQAiGLRpmEnw8Ils+QSj5Ly7AqdmyzbKoyQcutESqKdrWjlCc/OyiJVac0yHX2Uv2mMP3cK0UpAtdk4OS4GxfOaGq0n1DmR9axjCkvrd8YxFSOk0Bci1ToX7SWGk0Ya9qR/LMn1OdwbdLxjUx3/EzvVNtoXtPOg9O7KoPCfrEORu98z+nucaLe9Y8uzbnCathqbSHXbCrXtoNStHEbFFRAAIAoCEAAgCgIQACAKAhAAIAoCEAAgCgIQACCKRZuGnR3tuiybJVW1oVM6y2YWnKIaQ5npvwGKln+8VDncVhq2SgO2srBVBrFIWy7muELu2aZSztOBTuVNe/7xxgmdyiv3rUoCjHkq1DneMM6npn+8EGnYLZEiX8nb/v2mfX1MjRP+8cGESIee0Cdq3vG/vXXbYllw59ybPf+2x7v+VOofj03I/f53Z5V3bHVb5LI7597ROu4dW9bwp1K3xUrZlYFYXn1V+iO5rXMnZv3psdxYjv+nuAICAERBAAIAREEAAgBEQQACAERBAAIAREEAAgBEQQACAESxaOuAkskTLklnKeBoGoc8MeYdKpMFerhG3YaqzRipbkPVNRklN+mwDK6NsfYdSrVUKEf4U0nVPCViHiqZqAPKTuil6pOeGFd1Esb5lIjXQNnWdXJJy79t2hftMIzasFLUCaUDXXPT6Pqf3N5K/1h/YNRLtUXriY7RXkLMxfGuf46nxvw1QpXDHf/70/+K967KiYmWd2x5018HNJbp87SZ+uuE1jX/R26beeqAjhutb07hCggAEAUBCAAQBQEIABAFAQgAEAUBCAAQBQEIABDFok3DLrs9V6Zn5n4muU7pVO0arGgr2xuIdOl8TB9TIVJjVbsFa4n8MhVjOhvXYByTSP9W6bimMqylwij7tVPoxWMVbRFO/oJqeaF6WhjzL9oMqFYNJ8dVSYCYC9W+oyLORXUO1/crtlUp9KmRhq0kubWtaAMhtsqN/U4Nw485U7nw4y44DbsQNQ7/ky9zISZpxwAAWMwIQACAKAhAAIAoCEAAgCgIQACAKAhAAIAoCEAAgCgWbx1Qr+fK2fLejXqEpOmvySmN+gq1zH0h6oCKjq69GEyI5fONZ0DW3IiHU6pWDdVfHg1RX2G0KChEvYgas/Yrx40yFMWaCy0LbuWQBt6vqoux6spyo66sDHzuzDosVUJk1IaFttoQXQROHpI6nYbhNYGl+Lu9MOqA1Hh3lrrH0x3J/LU1Yw1/rc+F7Um534EoGjycj+vH45mLE/ncCve4AgIAREEAAgBEQQACAERBAAIAREEAAgBEQQACAJwbadjPP/+8e+CBB9zevXvdG2+84Z566il30003TY+XZem+8IUvuK9//evu8OHD7rrrrnOPPvqoe/e73z2v+0lXrnBp2p7liHXKc9lp+8fa+uGWLf94LpbAV2nWlf5yf5wfjOu0zXxMLXMvtmu5cEb2cBmY1ZwamZmpWjW+CE+5HUXW8++4cUI/742u/6CzvmozYDyYER6rSolW6erqXLNSx4dtIzW5EZg2bh5T2Fg9rt5mkvDnJinE/OdGGw7xwuvl/kn8375OpVbtGsbTvtx24KkjmcqNPPfQK6DJyUl31VVXue3bt886/sUvftE9/PDD7qtf/ap74YUX3MTEhNu0aZPrdrvzvSsAwHls3ldAH/3oR+vbbKqrn4ceesj9/u//vrvxxhvrn/3FX/yFW7NmjfvmN7/pfuM3fmP0IwYAnBfO6ndAr732mjt48KDbuHHj9M9Wrlzprr76ard79+5Zt+n1eu7o0aMzbgCA899ZDUBV8KlUVzynq/59auyttm3bVgepU7d169adzUMCACxS0bPgtm7d6o4cOTJ9279/f+xDAgCcawFo7dq19X8PHTo04+fVv0+NvVW73XYrVqyYcQMAnP/O6mrYl156aR1odu7c6d73vvfVP6u+06my4W677bZ57atcPuHKbJaUamNF60KkUpdNa1uxyvC4f7/Dcb3f/nJ/+mR/pU5R7a/w53XmHf92xZjOWy7FatiuYeQ8i1RSNZYM9WNV4+ZqzOp+R0hbVqnhjUn9eBon/OdTc7IMGqv32w1P4Zap1nKlbLlblzfDt1Xp1HLM2K9MtbbSsI2VqYOJl1apXlfVpmK8l/vPtaN98UbhnOuINOxlWU9um3ty0rtzTMOedwA6fvy4+4//+I8ZiQcvv/yyW716tVu/fr27++673R/90R/VdT9VQPr85z/vLr744hm1QgAAzDsAvfTSS+5Xf/VXp/+9ZcuW+r+bN292TzzxhPvd3/3dulboM5/5TF2I+ou/+IvumWeecZ2OjsIAgKVl3gHowx/+cF3v45MkifvDP/zD+gYAwKLNggMALE0EIABAFAQgAEAUBCAAwLlfB3Q2Va0RyuzMwysbOmbmY/7CgbyjWzmocVXr0xN1Pie3HWFJeXHIxbi/qCBdpZdRT0RxTJLqOqCi8M9FmY+w3Lyqg1igsgxzv0PRSmOFft7Trn/bxpR/u2xKz1PjRFj7iEoinlpVL2XW3Ij6Iqs1SCHGZcuR9kKdFNX7zMLUF6nxZKFqjwx98eQeHuhWDj3Ptr2+6q3yf7gCAgBEQQACAERBAAIAREEAAgBEQQACAERBAAIARLF407CbqSuzbF4tEyrDCdE2YUJvOxhLglKpBxNGGvZYWAqqlfJZjvuXPL9g5aRbKP2hSFcXqdZlqedJSY0U1VSkjqv7lanfIx7zcOCfp15XvPSm9HmaTWZB6d31tlNJUOsJVQ5gnafFLF1VTpe3yrAyhOYIKeeiXODk/ZZhr0njPFX7TUbpGzKCoSirODwQb17VqZrPniff7+sykFO4AgIAREEAAgBEQQACAERBAAIAREEAAgBEQQACAERBAAIARLFo64D6F3Rc0eic8fO8rWNmb6V/vG+0TVD1PPmZhzLnlgqqXsHctl0ELd/eE7U69bZiLDXqEVStz1Dcb6ORy/22xLiq86nHA8t1ihFKL6z7HDZFC4+m/7H2W/qkyNX5ZLw+slYaVAckTxirDkjU+cxlPJSq4UoaRr2OqudRdUDGfp3Yb2LVEInHo2rzTqRGsaFQGHVww2z2yRjol+s0roAAAFEQgAAAURCAAABREIAAAFEQgAAAURCAAABRLNo07O6qhms0zzy8YUenBfYuEKmKK/V9DpaJ9NYxkVc4x5TD2ZTGkvJOpHUmYip6vWZwewMr5VmlWuciRdtKw+60/HnAWaSl6lVKembMkzLIRauGjn5ZTrb9abW9jn7ehyLFO+mN8PeoOJ/Mc7wl5nEoTnKjpYJz4piMLXWdgkrRNvas2jGkLjgNW51PUyq9fkTDcvbX9MDfKWYGroAAAFEQgAAAURCAAABREIAAAFEQgAAAURCAAABRNBZzaJxthd1SL/Ls8rYY6xirzYp0UXOV21BW2qZKPy78aZnDvrEatkgHTa180MC0ZZU2XhmKVNLcSMNWKaoqXbqV6dTwdsOfT9pM9bZ5kQaNWendY+1+cKp7Pt4PSq8vxblWKcTjkedwvXP/vovMP1YOrLxlcczWy1mlSzfE82Oc44l8fZRy27IMO/9HW/Fdb9xIZj/fSs/Pz9h/0FEBADAiAhAAIAoCEAAgCgIQACAKAhAAIAoCEAAgCgIQACCKxVsH5DFbbdCMcVH+UhiPtrRqcgJqBsxx6y5F/UUpljwvRU1NTTzWwpiHrOmvg2g08+A6h6Fo5VComg5jvCVqeTpirJIl/sfayfQ69/3Ef8JNDZvB8zQh2la0G1Ny27GGf9tUnIyTQ38LiMqJgf/xnOjpbbt9/7bDJAs+T1XtklXXlIpan1TVCI3SNiTR2xbimNXYKBpGTZpvvJxjqxKugAAAURCAAABREIAAAFEQgAAAURCAAABREIAAAFEs3jTsKqswJLNQZDIaq+e7Mhd3KLY1Ey/VL1gbi/TiRKQtq+O12kuU2Qjp0iIdNFfzW20rU7hdsFRs2zTaMXSy4QhL1RfBLRdCW0S0xfFaqePqeAsXngafGfPUbvqPuS9aRAxVCwjn3GDgf3sbDowyhUCqLUI9rgaLNLgdg2w9McLrQ51r6nkfDOd433P7NQAAzi4CEAAgCgIQACAKAhAAIAoCEAAgCgIQACAKAhAAIIpFWwdUpIlLsvnntqci/zzt621LUWwiVzsvRki0N+uAxG57/v2mfavmRoy1jIMailYOA7EEfie8VqEhlsc/Oe6v52mIWh+rbqZljCupqKtR95un+u/Cpihos2qTVL3O0IW3w1A1RO3WMHjJf1XrMyh0Lc+xXts7Nmm0iBiIOqGhqE2StYRWnVCRhLeXEOdMYdQXKeq1Uxk2Zt/3cDC3thRcAQEAoiAAAQCiIAABAKIgAAEAoiAAAQCiIAABABZ/Gva2bdvc3/7t37of/vCHbmxszP3CL/yC+5M/+RN32WWXTf9Ot9t19957r9uxY4fr9Xpu06ZN7pFHHnFr1qyZ14GlRenSvJzXcv81kf2XGNumIr1YLnc+Sq8Ag8qqTVTK+dA6JpUmaaWShv1NUxqpmXnmT29NU72tOuKmSPNV7Qms9OJhqf9+K4xxn9TIzc9FWq0aq/TES149nkGuU55VmvZY05hj8fy0RMq5GrOOyUorl/uVLUf0PJVD1UIl0Xes3p7EbgvjOmOQ+s+JbqMZlFaeD+bWbmRer5Bdu3a522+/3e3Zs8c9++yzbjAYuI985CNucnJy+nfuuece9/TTT7snn3yy/v0DBw64m2++eT53AwBYAuZ1BfTMM8/M+PcTTzzh3vnOd7q9e/e6X/7lX3ZHjhxxjz32mPvGN77hrr/++vp3Hn/8cfee97ynDlrXXHPN2T16AMDS/A6oCjiV1atX1/+tAlF1VbRx48bp37n88svd+vXr3e7du2fdR/Ux3dGjR2fcAADnv+AAVBSFu/vuu911113nrrjiivpnBw8edK1Wy61atWrG71bf/1Rjvu+VVq5cOX1bt25d6CEBAJZCAKq+C/r+979fJxuMYuvWrfWV1Knb/v37R9ofAOA8Xoz0jjvucN/+9rfd888/7y655JLpn69du9b1+313+PDhGVdBhw4dqsdm02636xsAYGmZVwAqy9Ldeeed7qmnnnLPPfecu/TSS2eMb9iwwTWbTbdz5053yy231D/bt2+fe/3119211147rwOrUoxnW0Q60ZmXI0nFSs6pyiQ1Mg7lAsUjbKtSL60MYDWPavHuisr0LVUWqrVSsEhRLRpGCr2YqKZY0bdjrHatVp7uD/QfTlaatk9hpMHLFaKNNODe0P+S74tVnkeRGynPMoW7MQhOoVcrjhet8FWr81ykwVtzqMojihHKOdT7hLFp7lnRei7nU+J53Vnp6EEBqPrYrcpw+9a3vuWWL18+/b1O9d1NVRdU/ffWW291W7ZsqRMTVqxYUQesKviQAQcACA5Ajz76aP3fD3/4wzN+XqVa/9Zv/Vb9/x988EGXpml9BXR6ISoAACN9BGfpdDpu+/bt9Q0AAB/WggMAREEAAgBEQQACAERBAAIAnDuFqG+HdFDOuiy9WLG/1phKwsNtuSDdCxbMKHVAqoVEIQuXqhoiVcsgNjTqQcoR6iB89QhWS4V2OgyuL1JjcxkPqdWpHOv664+menr5/OHAv+9iIE4a65wQ7TJ6ff14um1R69Pyjy1r9YPrpay2FUXYU2cX3YhzPLFaqKjXnZj/MiuD67Cm9BG5fmP2N+TixNwmkCsgAEAUBCAAQBQEIABAFAQgAEAUBCAAQBQEIABAFOdcGrZ1wO0jRViK9ghpzUXTWG6+Edi+wLhf2VJBZxfL1HHrmFSqtTqmOSwluCBkirZot1DJxLYtY9t+Hvby6hpp2EePjXnHiqM6DTvt+U+odBiW5mudM70x/XgG4/7xE62Wd+x4S7fDaKT+EzUVY5W+eA4KkUpdilYNJzf2DyV9/T6SifFSvKDLhpGGLU6ZwkgNL9LZ56mYIg0bALCIEYAAAFEQgAAAURCAAABREIAAAFEQgAAAURCAAABRLNo6oCQvZ13iPbOKSRJ/TM306u2uyMLqHPKWC8+zb4TXEI1UyyNbT+hjSsTGgR0IzI1F94haJupUVD1IU7RqqKRivGnUAaltSzHHg6F+8oqufzw7of+mTEUtiaodK40noGwEvrCMupp+w7/jQUu/faVNURPY0M9dIep5CtHKoVQtLarzWIynAz3HaS8Jeu2o+TVr94ZpWJ2iOEdPxxUQACAKAhAAIAoCEAAgCgIQACAKAhAAIAoCEAAgikWbhp0WpUvzcl7pq5Ws50+9TKx0RJESLVMZjdRwtcx9kYUvcz8c9+831yvV22naEcyWdn9Klul06VbDn0PcyQbesXbqH6ukIr11ykzh9m/bF6nJQ5HmWzNeAwvBTK/Pw9oI1PvOs6BWAmWm5ylv+bfNW8UIZQr+oUS0uzDT4Af6kGSLFXFOqDTrSqHm2DjXvMPGc34KV0AAgCgIQACAKAhAAIAoCEAAgCgIQACAKAhAAIAoFm0ativKk7e3SGb52VxTGa3VZoumSGtu+WO1kY0rV8OWS1pX96v+RBhl5WmVSmo8HpXWmYhVk5O+tbKuPx23l8lJdEezjnesEKmkx/r+7eptxfNzYqCP6XjXnws/1fVvO5jUy6unx8Vq2N1kQVbDtqhsXSvlX2VTe1dbrvcbvpJ80VyYv73V/Fay3sKkYZdJ+PtEKjYujRIT7/NjPJbp+57brwEAcHYRgAAAURCAAABREIAAAFEQgAAAURCAAABREIAAAFEs3jogj9laNMwY7/qLVJJcF7jk482g2pfheBpcB5TrUhKXd0SOvrjbUWo6jDIgWTeQiOfHqsMqJ/2nY3FMF5NMjvlrZybL5d6xQ3kSXlcm2mxUErFtJuaiadWDqDos43lX54VV/xVK1eOY42nY+W/WJon9jsKaQ3U+JSPU9am5sOZfM85xz/2WxmvjFK6AAABREIAAAFEQgAAAURCAAABREIAAAFEQgAAAUZxzadhmO4auP880yUX+6gis5eYLsd58KlpA1NuKtOZMPZyePia1lL2Z3qrGRTsGKzU8HZRBc1iPiw4G6n6zbnh7D7NthThXE7HMvVxaf8TnTrbhWKD2HtbrQ7dcCNtuVPI5GKH1gZrjwmgvIedihHdyud8ybJ6scoBTuAICAERBAAIAREEAAgBEQQACAERBAAIAREEAAgBEQQACAERxztUBlaLOpKZy6YskuG4j7fmLbppGi4is64/zjaksuB3DKFR7iXQ4yrrw/qHmcV0c0DzcDb9b9byrc8Y4n3TNjbVtGrTfomHUhrX8+y2surJGYHsPo4ROtuGwzidVT6VqhMz5F2Pm8x5WrzNKDV0uatnq8bbYrypcMurVRqkv8j2eco4ll1wBAQCiIAABAKIgAAEAoiAAAQCiIAABAKIgAAEAFn8a9qOPPlrf/vM//7P+98///M+7++67z330ox+t/93tdt29997rduzY4Xq9ntu0aZN75JFH3Jo1a+Z9YNVS97Mtd6/SPU/+wghpyyqVtBQpxEb2cEOsad7o6KegaGaBqcf6mFRaedoz1lIXaZ1J4R9M/+eo3G1+6MfesXLQd6GSpj+/NZ0Y09uOifGxjty2bDf9Y5l4Xts6Nb9o+8+Zwth22Ak7n6xU6nRQBJ1r9Xg/D0pld1Yatkw519vmItVd7ddMgzdaLijqmEvVN8EoPymL4Axu7yXMgrRjuOSSS9z999/v9u7d61566SV3/fXXuxtvvNG9+uqr9fg999zjnn76affkk0+6Xbt2uQMHDribb755PncBAFgi5nUF9PGPf3zGv//4j/+4viLas2dPHZwee+wx941vfKMOTJXHH3/cvec976nHr7nmmrN75ACApfkdUJ7n9Udtk5OT7tprr62vigaDgdu4ceP071x++eVu/fr1bvfu3d79VB/VHT16dMYNAHD+m3cAeuWVV9yyZctcu912n/3sZ91TTz3l3vve97qDBw+6VqvlVq1aNeP3q+9/qjGfbdu2uZUrV07f1q1bF/ZIAADndwC67LLL3Msvv+xeeOEFd9ttt7nNmze7H/zgB8EHsHXrVnfkyJHp2/79+4P3BQA4jxcjra5y3vWud9X/f8OGDe7FF190X/7yl90nPvEJ1+/33eHDh2dcBR06dMitXbvWu7/qSqq6AQCWlpFXwy6Kov4epwpGzWbT7dy5091yyy312L59+9zrr79ef0c0X1l36LLGcF4rVv/0gOZ9X9P7VqmMIs00mdIpwkl/4B1LRTpupWyItNmxVlCqbn1MuXg8A2vpY5XeKi6qW/605Eq6bMK/377e1uV5UCp1MjEud1uOtYPmvx43nlv/QRmpuuI0tV4fKl26FEs1m687xUg9LlWpgUo9tlYNVyuON/WHP2rfeSu8/EFJrAoT9dYmxlJ1wtRZ2knYfQpz3a4x34/LqpqfKrHg2LFjdcbbc8895/7+7/++/v7m1ltvdVu2bHGrV692K1ascHfeeWcdfMiAAwCMFIDefPNN95u/+ZvujTfeqAPOlVdeWQefX/u1X6vHH3zwQZemaX0FdHohKgAAIwWgqs5H6XQ6bvv27fUNAACFteAAAFEQgAAAURCAAABREIAAAOdmHdBCqeoOZq09sOoRQmtUjDog1QZC1fnU93tC9GswloVPjGP2bidqKyxW/UopaihUTYdz/jqfSiqW3k+HeXD9V9kW9VLjugg6H2sGtz5QNSyj1NWotgmyRYfVDkCcalb7AqdaFBj1OqrVg3o8VmuD4Foe634b4fUvoXU1Vp2QKmE0ybqysF3OdTuugAAAURCAAABREIAAAFEQgAAAURCAAABREIAAAFEs2jTswbKmKxvNeS9ZrlIKU9GCoB7v5UGRulyu04uTCX87gLKln4K84x8fjvtThPMx/bdF3kpHSFF1QamxSdGR+02Hy/2DVoeCNHBZfuMVUIpxc1srddm7nTEenmHvUlExkA5EqYFV/SCy5LO+1SJC7dg/VBrTq87johX+PqLfY8LnyRSYal00jdT8Rvi56Ht+5nqOcgUEAIiCAAQAiIIABACIggAEAIiCAAQAiIIABACIggAEAIhiEdcBNVzZPLuHl/V1HZC/qka3gShFPU49LtoMDCf0YxyOi23b/vz+YUfn/g8n/OMDXdakawOS8LoYWcvTLIPrdax6EUm2KDCOSY2r/WbWY1VtQ/SDbRz333FjUtRwWfUt4qWVTRnH1B2ll0BYHVDeHqFeaugfK0UtlbVtMkorhyS8Xk2OB752zPqhn+IKCAAQBQEIABAFAQgAEAUBCAAQBQEIABAFAQgAEMUiTsNOXTFbywBrWXiRLl00w9exT5v+WF02dBwvGiJdesxIlxZtFXqr/Nv2V8rdurzjn6fcWKpeLs1fjpBeLJ6e0krDboocVZW23NC5r4lKpTZSVJMkbBLNFfDFtsOBPseHhb9kIMn995z29YNN+2pwhJRdNf3me0FgCwijhUTWC0uzro8pD085L1UZQzlCerd6zVrPna8dA2nYAIDFjAAEAIiCAAQAiIIABACIggAEAIiCAAQAiGLRpmH3VyQum201WyOlMBPplUXfSin0p7Bm7TQozbqSN8NW7LXStLsX+fMne+80li8W6ceJSFuulENxzEPxN42Rhp2IVOq0pR9Ps+kf77QGQWP1flPjhFsAhbF897Dwz/FkT+fQH+/7t81VqrWV8izOCTOVV6QXq7KKUUoyUnUOG6nWavVue9Vwsap+po+pkHnn4c+dYq0k73tuScMGACxqBCAAQBQEIABAFAQgAEAUBCAAQBQEIABAFAQgAEAUi7YOaNh2rmzPf7tcJK5bOfoDVRtQhLURqMfFLBfGtmL1fDdYLuoclg2C2wwYqf/OGe0afJotvVZ9uzkMGqt0Gv7xZuZ/4jNjrfrUWvN/AQyMk+LEQLRUMI43afkfbzEW3npC1X0kxQitHOQdG8+NGE6HelvVViEdjHBOiB4dhfFurGoG1fuE+f6Unf1tS+O99hSugAAAURCAAABREIAAAFEQgAAAURCAAABREIAAAFEs2jTsvFPl3p7ddFBriXCVwaqWJS9FSvOoYV6lQRbL/bmiK5Z1g5f8L4y02UbmT+VtNfz5l8vaYo1759wF7RPesU6m07Abqf9+C/HED42TwmqNsBC6qn9H1Wpj6H/ZZkYadibacAw7/jnMMz1PZeofz7p6Due6dP+ZjPRukWqd6ioFve0w/LGodOnCausiyh+KpmiHYZV6qPcYI0L49l3ol+s0roAAAFEQgAAAURCAAABREIAAAFEQgAAAURCAAABREIAAAFEs2jqg3oWFS8dmqVmwlsdXqfRmDZFav11saNUBjbCivNq2uawf3L6gHKG+pS1aH6ix5U1dBzTe8BdntFXxhdE2YVj6x4rcmAcxnBpPXmgrB6v2SLWXSFPdXiITdVp503+Sl8Y8lZmY46aeh6Id1kIlMWtNVPGePqZhR9TJZeEtFWQdUCu8DkjW+oxSyrbAZXBcAQEAoiAAAQCiIAABAKIgAAEAoiAAAQCiIAABAM69NOz777/fbd261d11113uoYceqn/W7Xbdvffe63bs2OF6vZ7btGmTe+SRR9yaNWvmte/0oq5Lx0c5uln2OcJS9Q2RvmrtV43nRsqtSpfuiFTrlkjVrTREum5mpPK2RWuE8UY/aKwylg2C2wwoZqq1C0u1Vi0gRjlmq0VEI/E/P03RKqPeVrVjEGNWOwb1p2ypu0s41X1CdHlwyUgpwnpjmU7dEdsZj7Voi7FMb6tOC3mqGadheDuM0QXf9Ysvvui+9rWvuSuvvHLGz++55x739NNPuyeffNLt2rXLHThwwN18881n41gBAOeRoAB0/Phx98lPftJ9/etfdxdccMH0z48cOeIee+wx96Uvfcldf/31bsOGDe7xxx93//RP/+T27NlzNo8bALAUA9Dtt9/uPvaxj7mNGzfO+PnevXvdYDCY8fPLL7/crV+/3u3evXvWfVUf0x09enTGDQBw/pv3d0DVdzvf+9736o/g3urgwYOu1Wq5VatWzfh59f1PNTabbdu2uT/4gz+Y72EAAJbSFdD+/fvrhIO//Mu/dJ2O+CZuHqokhuqju1O36j4AAOe/eQWg6iO2N998073//e93jUajvlWJBg8//HD9/6srnX6/7w4fPjxju0OHDrm1a9fOus92u+1WrFgx4wYAOP/N6yO4G264wb3yyiszfvapT32q/p7n937v99y6detcs9l0O3fudLfccks9vm/fPvf666+7a6+9dl4H1u4MXdbxp+X6JCIfUaUeV5oi1VqlNY+Shj0s0oVZlVqkSo+6GnNLpB+rsbaRtrws06tlK7n4W6ohllSeaPSCn7umWqrZMJDLF2uHxTFZz516faiVtAuRol0pG2IlbStdWrwECrWtMYVFa4RV6AOVxjtq3vLfcWItqq/KCdTTo586nUKfBnYQsDoEhASg5cuXuyuuuGLGzyYmJtyFF144/fNbb73Vbdmyxa1evbq+mrnzzjvr4HPNNdfM564AAOe5s94P6MEHH3RpmtZXQKcXogIAcFYD0HPPPTfj31Vywvbt2+sbAAA+rAUHAIiCAAQAiIIABACIggAEADg/suDOlrHWwGWtdF51DFYrAWt5/LGmv+6oadSwLBRVh9IR7Qs6Rh1QN/c/9X25Fn1VEzUMalHQTofB7RhGqatR9TqdVN9nW4xbdUCFKKI4IYpUhta6/IHtO+w6IFGjkunXTinGR1nuX7U3UPdp3u9IrRzEfVq1SU0x/wN9UGlPbDtMgh+qmidvnY/xeOda5sYVEAAgCgIQACAKAhAAIAoCEAAgCgIQACAKAhAAIIpFm4Y90eq5Rvvstj7IEr0uuUpdVqnH1jEp1vL5imp9YKU0p2IuGoWep5ZIp1YtF1SK9lyen1Bqv7mRI9wTecA9J3KEjX2rtPGBcUwq1bowTsVCtP8ocpWP64KVIvW4MhwT22bhadgq/9huMyAG1f0aL+dSvVeU+qDSfvK2P1YznTod7dKGKyAAQBQEIABAFAQgAEAUBCAAQBQEIABAFAQgAEAUBCAAQBSLtg5oVXvKNdtntwVCQ7RqqMdVbYyoYWka9SuqTsiqA1I1IarmxmofocbzMrwtgmp5kbnwOp/UKERJRWuEQhRJ9IzWE6peZyr3t1Sw7lc9Hqsdhtqv1Y4hF7U+aqxUNUIj1gHl42JbVXMzwp/Psh6n0hCtKVr+87gsjEKgXDx3A72prPVRd2vVJqmWC1YNkWffcy1v5AoIABAFAQgAEAUBCAAQBQEIABAFAQgAEAUBCAAQxaJNw25luWtm+bzTlkdpb6DSWwuRDl1YLSJEyq3VyqEt0otVKnVDbGdJI/1dIlsjWKnugf0CrPkvRFqt1TZhWFhr2c9uKtdtHoaipcIoEpXma27s/42yoZ+7MhN3LLZNRKq0mRJtdXIQ95uK+y1EmvXJuxXPXcM4F5v+fSdi/kfqclKEnTNzvU+ugAAAURCAAABREIAAAFEQgAAAURCAAABREIAAAFEs2jTsUCoNOzQttpLKlbL1flsiXbqdDuW27WwYlGpt7Vet8lyEZTSbcuPvHXVMFrUatkzRHmElc4tK01Zp/f1cz4O14rUi03XVWGakUquUdPPPXLGCesd/HrdawwVZ+bsiF5dWKedGar5Lw9PV847YbV+svD7Q54t6CYRWcyT6qZnGFRAAIAoCEAAgCgIQACAKAhAAIAoCEAAgCgIQACAKAhAAIIpzrg7IarfQFUvZ58Yy9lmqan3899sxkt7HsoEY68ttrXqe8+m5k20G0sV3zKqWx6o764sxdQ6f3G/4ZKSiDiVr+Is+EnH+1/vNwuulVF1Nu+N/7Uy09WunN2gEjVltOIoR2jyUql1DadR3ifKwQp0yVuuJYoRtfZvNsX6IKyAAQBQEIABAFAQgAEAUBCAAQBQEIABAFAQgAEAU51wattVSoZeL1Muhfrjthj/lOU1UKrV/rPKO1jHvWFO0aqjHxXroXZF72SsW31NrpS3LWSzC/5RSqdRWC4i+mMeecS6Gplqrc9hKw7ZaNaSi1KDVsibZL7xBhG55MS5SrZc1dRr28bQVlPpd6Yv3isHA/7yWKkW73lidqE4qZbsMsVv/NNgC25EURpuTU7gCAgBEQQACAERBAAIAREEAAgBEQQACAERBAAIAREEAAgBEsfiKRQyFUXGgWi7kRo2E1a7Bp53plgnjouWCqvOppCKfPhd1Nb1z76mVdUK5UY+QlmVgSwV9TgxELY9V1xR6v9Z5OMrjSQLrcVT9UCUTbR4s6n47ojZP1e1ZdViDTNdwDQv1eMPrgBLVjsFQijlWY6PUaJVWmw3Pzst8bucDV0AAgCgIQACAKAhAAIAoCEAAgCgIQACAKBZdqlT502ymweTsmWMDnfjihj2RXZQbmS+ZPyNtMPBnsvVLvSpvd+Bf5zk3s+D82SS9PA0aqwzLIjADKI7CWDW8EPOoMsPUPFT64m4Hhc70GQz948NczP9A7zcfiszIgZ6n0NWWSyMLzo2QBVc6MU/idTdo6lXoh+JlmVtzLOax6ImxKT3/yVT4W26p5lgMJdYK3QuQBVd0uye3F9mp9eal9Rtvsx/96Edu3bp1sQ8DADCi/fv3u0suueTcCUBFUbgDBw645cuXuyRJ3NGjR+uAVD2QFStWxD68RYt5mhvmaW6Yp7lhnmZXhZVjx465iy++2KVpeu58BFcd7GwRs3pyeYJtzNPcME9zwzzNDfN0ppUrVzoLSQgAgCgIQACAKBZ9AGq32+4LX/hC/V/4MU9zwzzNDfM0N8zTaBZdEgIAYGlY9FdAAIDzEwEIABAFAQgAEAUBCAAQxaIPQNu3b3c/93M/5zqdjrv66qvdP//zP7ul7Pnnn3cf//jH6wrjaqWIb37zmzPGq5yS++67z/3Mz/yMGxsbcxs3bnT//u//7paSbdu2uQ9+8IP1ahrvfOc73U033eT27ds343e63a67/fbb3YUXXuiWLVvmbrnlFnfo0CG3lDz66KPuyiuvnC6ivPbaa93f/d3fTY8zR7O7//7769fe3XffPf0z5uo8DEB//dd/7bZs2VKnOX7ve99zV111ldu0aZN788033VI1OTlZz0MVmGfzxS9+0T388MPuq1/9qnvhhRfcxMREPWfVC2Sp2LVrV/1msGfPHvfss8+6wWDgPvKRj9Rzd8o999zjnn76affkk0/Wv18t/3TzzTe7paRacaR6M927d6976aWX3PXXX+9uvPFG9+qrr9bjzNGZXnzxRfe1r32tDtynY64ClYvYhz70ofL222+f/nee5+XFF19cbtu2LepxLRbV0/fUU09N/7soinLt2rXlAw88MP2zw4cPl+12u/yrv/qrcql6880367natWvX9Jw0m83yySefnP6df/u3f6t/Z/fu3eVSdsEFF5R/9md/xhzN4tixY+W73/3u8tlnny1/5Vd+pbzrrrvqnzNX4RbtFVC/36//Mqs+Qjp9nbjq37t37456bIvVa6+95g4ePDhjzqr1mKqPLpfynB05cqT+7+rVq+v/VudVdVV0+jxdfvnlbv369Ut2nvI8dzt27KivEquP4pijM1VX1R/72MdmzEmFuQq36BYjPeUnP/lJ/aJYs2bNjJ9X//7hD38Y7bgWsyr4VGabs1NjS021unr1Wf11113nrrjiivpn1Vy0Wi23atUqt9Tn6ZVXXqkDTvURbfXdxVNPPeXe+973updffpk5Ok0VnKuvAaqP4N6K8+k8DEDA2fqr9fvf/777x3/8x9iHsihddtlldbCprhL/5m/+xm3evLn+DgP/p2q1cNddd9XfJ1bJUDh7Fu1HcBdddJHLsuyMTJLq32vXro12XIvZqXlhzk6644473Le//W333e9+d0aLj2ouqo94Dx8+7Jb6PFV/ub/rXe9yGzZsqLMHqwSXL3/5y8zRWz5iqxKf3v/+97tGo1HfqiBdJftU/7+60mGuzrMAVL0wqhfFzp07Z3ycUv27+sgAZ7r00kvrE/70OasaZlXZcEtpzqr8jCr4VB8nfec736nn5XTVedVsNmfMU5Wm/frrry+peZpN9Rrr9XrM0WluuOGG+qPK6krx1O0DH/iA++QnPzn9/5mrQOUitmPHjjqD64knnih/8IMflJ/5zGfKVatWlQcPHiyXqioT51/+5V/qW/X0felLX6r//3/913/V4/fff389R9/61rfKf/3Xfy1vvPHG8tJLLy2npqbKpeK2224rV65cWT733HPlG2+8MX07ceLE9O989rOfLdevX19+5zvfKV966aXy2muvrW9Lyec+97k6M/C1116rz5Xq30mSlP/wD/9QjzNHfqdnwVWYqzCLOgBVvvKVr9RPbKvVqtOy9+zZUy5l3/3ud+vA89bb5s2bp1OxP//5z5dr1qypg/cNN9xQ7tu3r1xKZpuf6vb4449P/04VkH/nd36nTjseHx8vf/3Xf70OUkvJb//2b5c/+7M/W7+23vGOd9TnyqngU2GO5h6AmKswtGMAAESxaL8DAgCc3whAAIAoCEAAgCgIQACAKAhAAIAoCEAAgCgIQACAKAhAAIAoCEAAgCgIQACAKAhAAIAoCEAAABfD/weYhE5nUTZVfwAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1. 0.]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1. 0.]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1. 0.]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0. 1.]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0. 1.]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for imgs,labels in train_batches:\n",
    "    for i in range(10):\n",
    "        print(labels[i])\n",
    "        img = cv2.resize(imgs[i],(50,50))  # resizing image for plotting\n",
    "        plt.imshow(img) # plotting image\n",
    "        plt.show()\n",
    "    break"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0. 1.]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1. 0.]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0. 1.]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0. 1.]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1. 0.]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1. 0.]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1. 0.]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAaAAAAGfCAYAAAAZGgYhAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjEsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvc2/+5QAAAAlwSFlzAAAPYQAAD2EBqD+naQAANCpJREFUeJzt3X+MXNV5//Fzf8zM/rB3F0NjB2G3SIkgKYIoTgIuVZuCGyuKEBT+SKVIdVPUKBQQYPRtY6khatXKNEiBkBoSpRRUqakrV3UiIpUWOcGoqk3BKSohjdXqS4v7NTaljX+td2fur6/udbz14jmfs3vGmzN43y9pMLtn7507Z+7Ms3f2ec4TVVVVGQAAfsLin/QdAgBQIwABAIIgAAEAgiAAAQCCIAABAIIgAAEAgiAAAQCCIAABAIIgAAEAgiAAAQCCSJdqx9u3bzcPPfSQOXz4sLnmmmvMV77yFfORj3zEuV1ZlubQoUNm5cqVJoqipTo8AMASqVd4O3HihLn00ktNHIvrnGoJ7Nixo2q329Wf/umfVq+++mr1m7/5m9XU1FR15MgR57YHDx6s16bjxo0bN27mnX2r38+VqP7P+Y5+1157rfnwhz9s/viP/3juqmbt2rXm7rvvNp/73OfktseOHTNTU1Pm+g/9H5OmnXPGy7b+1LA3Yb+oy8f0tkXLfsVVimvFqJS7NZW4kOtN6Ku8fNwMnTIRg2nlNYe1Smxbxf5zrI6pShynf8v+5MbtQm4axX4vrSTV+1051rOOjbe7ctsVbfu2Y6l9bCTO5H7HE/v4SGLfb3O/Ynw0FsfrOKYQsipxjMfe2+ZifCo9ZR1bkx6V+52M7dtOxPp8SqP+5+r0ydLctOGQOXr0qJmcnLRvb86zXq9n9u/fb7Zu3Tr3vfoSbOPGjWbv3r3n/Hy3221uZ9SXbc2BpR2TpiPn/HyZ6neismV/SFXL8S7Wtr+LRUsUgJKODkDVuTE4ODUXKoiYpQxAcYAA1AkTgJIx+znT53e2eVptMSaen3aiz9O2eO/sOOZYjY/EfmOhJI4gosYTZ/Cyj4+k9idvrKX3Ox7bx8fVx2f1ORPp58D1Z5TznoTw1ltvmaIozOrVq+d9v/66/nvQ223btq2JkGdu9ZUSAODCFzwLrr5Sqj92O3M7ePBg6EMCAPwEnPeP4C655BKTJIk5cuTIvO/XX69Zs+acn+90Os0NALC8nPcA1G63zfr1683u3bvNLbfcMpeEUH991113LXg/VRo1t7crHX/HKcVn1WqsuU81XPn/DUjtNs4d2zrGve50QOqj96oUf0cr9OfFVW7f1vHxuPwbUKn+pOL6u5TYcVk4Jtn19yXbfh3n+Elxt4WY/9Pj9n3nKsFH/O2olooXQex4gbQsf8iuzZqW8VWo587xAlHHpMYKx37VXLSMP3VMI5FO1lAJJmq/p8fLRX3/J1IHtGXLFrN582bzoQ99qKn9eeSRR8z09LT59Kc/vRR3BwB4B1qSAPTJT37S/Nd//Zd54IEHmsSDD3zgA+aZZ545JzEBALB8LdlKCPXHbYv5yA0AsLwEz4IDACxPBCAAQBAEIABAEAQgAMCFlYQwqGw0NVWfxalKsV5bLR+1jxeOWga1rFEs0uHjzLEekkiJT2f0MVVqLSW1yvkwdrJwrAuljrlynKlqodNSLDIr165bwBp03o9H1DVVLX1Msx37g50d0Sf5yVF7zceKMfu22bijEOvcZRvnlAOcjF3xxJ6MOt7bqrHaisS+COdEOuu9aKusL4r1tmPGvjDrxclJ69hUYl9stNk2tr8JjTnrgCzfjxdWB8QVEAAgCAIQACAIAhAAIAgCEAAgCAIQACAIAhAAIIihTcOu0637pc+qlNpaFfunfKqMwzi3p8YmPf807MqRhq3aQKgU4UHSh13U45GrsFf+afCOrFmZni9TtPu0/FgoZyt21SJCrL3frw3J2Qr1WFVv7LpMYdw+GT8SadjdTD8BM+P2BzTesqcPnx63pzzHjpbPvlyp4bmYx1SkGCcDtJ4YGaBtwsrE/kYyFdvTxpttRY3JmKN0ohXFA70PcwUEAAiCAAQACIIABAAIggAEAAiCAAQACIIABAAIggAEAAhiaOuAbByrg5uka68bcKx2Lmt94nyAdgyVfTxyFOyo+1VcK+DLu42Wpl5KtbRo9lvat3XVFZSqvUEi6mYcXQbkXDhKVNT9Fh2/Oh/Xft31UvYHXIzYx2an9I7/38SodSxu6yc+bdnHk6T0GquNd+z1Rys79tqjWjuxH9Nobn/yEuNfB+Si6oTGI/tjHY/0m4iq9RmLWo7H0/+cyV1Fcj/GFRAAIAgCEAAgCAIQACAIAhAAIAgCEAAgCAIQACCIoU3DrrMV+630rVJ1FzKuqBRilT05yH2aUqcrxplIyR3gbiuVtuxIeVZp51Gh0uAdj1VsWzmWhVctDFRKuuuxym4AIr3elS6dj8b+adiyDYf/tio1PJnVv6v2ZuzpumVbv81kbfs89lpijlv6fMpWunLsjVdbhVSMxa52DKr1QazbVqh9t0Sqtesqw9ZS4fSYnkPbeMvxel3osQEAsCQIQACAIAhAAIAgCEAAgCAIQACAIAhAAIAghjYNO8lKk1QLW1F1ngFSomXK7QArTxuRGutMK1creMtVtgc8ZnVMYt8qDdtFpRCrlOZaKdKwBxGJ1PF0xn9l46QrVqXuxAOshj1AGnZP7Lfl2K9Iuy1GHM9dx37O5KMqNV+fa1nP/vZ2YkbknNfHXNonqlvY99tzLq/uv1J2S6Raj8f21b2nxFizrXifbRm/czxb4Hs3V0AAgCAIQACAIAhAAIAgCEAAgCAIQACAIAhAAIAgCEAAgCCGtw7oVG6SNF98rY5jiXy5aRL71aG4lsAXfRMix7by8YgaIseq8Ebe69KU1DgN0jbB+5gdp4tqERHPqiItXRMViVqSuOuoA0o9z1PHeNGOB6izEq0n9DSZQrQcUU9PIV6vtXLGPsezjuc9y+z1PDM9e+uJylFg1xbtGGLHm1ssZmNM1gGdkvsdiezjcayfvBHT/42mSx0QAGCYEYAAAEEQgAAAQRCAAABBEIAAAEEQgAAAQQxtGnbrRzMmTfqk8oll311pplWil0qvOvbx0vgvs66OuXSkF7uWwfel0mrVkv2DSLo6zVS1N1BtEWpVKR7PEqWVl2398omK0qt9gbNFR8+/DYRKZ49y+znu6Khg4tx+0mQr9MZZoc5FVf7gSMPOxetOpH4346n9uc1a9ue1FOehy6xIza/NdOzp30rb0eYhq0TbinhGbtux7PtkTho2AGCIEYAAAEEQgAAAQRCAAABBEIAAAEEQgAAAQRCAAABBDG0dUHT0hIn6LTHuqOWJ2mKpdEcefZl07Pt1LP2uqFoG1zL3RUdsq2qe4gHqgBwlT751NZ0TujYgmfVri+CqnVE1N4O0nlBtERrieVdz6KoDilVNVOmovxC1M+p+nbVJooYocpxQai7keep494pEHVDcdRxTKs4n8R4043hxvBXb95uX+nzqlfYH3BL9V0YcLRUKcR0ynbTltiNR/31PFwurVeMKCAAQBAEIABAEAQgAEAQBCAAQBAEIABAEAQgA8M5Iw37++efNQw89ZPbv32/eeOMNs2vXLnPLLbfMjVdVZb7whS+Yr3/96+bo0aPm+uuvN48//rh573vfu6j7qbKsb/pyVOl0UJMO0DbBk2oB0YyLVNKy7ViqfjT2SrV2pWHL8SX6tSQXKeXN3a4Qqby53rc6L6LcPpb0dNqyK/1YKVv2iSzb4nl1pIZHjpT0pUgNV+dwrejYH0/haCkiz8UB0uQjNU2uKRTtGiyZx41yxJFK3RMp3C39djyd2lOij2aj1rFjbftYrSXaNRSOk9HW6uHUUqVhT09Pm2uuucZs37697/gXv/hF8+ijj5qvfvWr5oUXXjDj4+Nm06ZNZnZWFHkAAJadRV8BffzjH29u/dRXP4888oj53d/9XXPzzTc33/uzP/szs3r1avPNb37T/Oqv/urgRwwAuCCc1w9bXnvtNXP48GGzcePGue9NTk6aa6+91uzdu7fvNt1u1xw/fnzeDQBw4TuvAagOPrX6iuds9ddnxt5u27ZtTZA6c1u7du35PCQAwJAKngW3detWc+zYsbnbwYMHQx8SAOCdFoDWrFnT/HvkyJF536+/PjP2dp1Ox0xMTMy7AQAufOd1NezLL7+8CTS7d+82H/jAB5rv1X/TqbPh7rjjjsXtLMuM6bOKsSt7Mqr06q2+ZDaiI1VUrVpdONKwixGzJGnYg5CrSyvisZxmP+jYlYYt0qXTGf80bFP6p9+rVOtcpNeXjrTlENQ5XFMLXpd6EXo5j74rr5/eWIy5Fg0X42LhaVN09QuvyOwT1c30RJ1K7anNJzP7Sv4/ysbkfmMxUZljJXNbCvds7njB+gagkydPmn/7t3+bl3jw8ssvm1WrVpl169aZe++91/zBH/xBU/dTB6TPf/7z5tJLL51XKwQAwKID0EsvvWR+6Zd+ae7rLVu2NP9u3rzZPPXUU+a3f/u3m1qhz3zmM00h6s///M+bZ555xoyMOH/9BQAsI4sOQB/96Eebeh+bKIrM7//+7zc3AACGNgsOALA8EYAAAEEQgAAAQRCAAADv/Dqg8ypJjImS/t9XYrHMvWPbKlXL5ydeS9E346LmIxtztGMYF60cxLNXOZ5ZufT+Ev1aErmKOtT4AJsmPftYOq0nKulVA9Qmec5x5b9fF1mvM0gtj2o54jgX1b4Le3mLKR0lf3IaXfVFqsZOzX+in7xogNdWUdo3nsntk/g/vXG531KcjKdECwhVB9St6zgXgCsgAEAQBCAAQBAEIABAEAQgAEAQBCAAQBAEIABAEMObhl2nU/dJqY5EmnWtSsS4a/n81G+J/GJEH1M2at82F2Onx/3SUItO5Z2OW6Wuphd+ypbjmNqVd3qrSUVubE+0eZjRqfnJjP35SU/p5y7p2sdikRoeOzJY41zMhWOaStH+Q6U8F66U5wHeSVQKfTFSeZ9Pg6hK+0GpbiSu104UqXU0K31MYqJkGnZXt2Poifz7lYXOv+8k/WsRepk4wc/CFRAAIAgCEAAgCAIQACAIAhAAIAgCEAAgCAIQACAIAhAAIIjhrQO6aNKY5NzChKqlD7kcteetF2Kslo/b952Nx961PKqlQm9Cbmp6U6IOQtTNyJqaph1A5b9Uvai5iUQdRKRqderdpoV9W9cxCYVqpeGo2yhFqw1X/VfStR90Mht51QjVokLUqLjqgBLf2jDXfv3PJ9Waouz4nf9OrpYX9lPRGDXmmn9RX5RljlYzS6QU9UW5aAFRayf9JyPLFhZauAICAARBAAIABEEAAgAEQQACAARBAAIABEEAAgAEMbRp2PmqcWPSkXO+X7laKnQS/7YJYyLVWmybjcvdmmyFSL2c1Hmb+aTI+Wzb05rjtsoVNSYeIK251e6/BHtttGPPIU4TnYbdsqR01rJCp6hmeeK1bd7S50Qu9luK1PzmfmfF/Yo2ECp9uxbZp99Eeoo11WYg8W+poNKsG6IkQKVhV63S/6AGmSe1X9emuXiPiRPvdgxLlYadqbz9+jVtOeFyce6fjSsgAEAQBCAAQBAEIABAEAQgAEAQBCAAQBAEIABAEEObhp1NtEyVnrt6dZVG3qnUaqyWj9nHihH7/RZtuVtTjKgxnYYdjdhTkxORat3uZHK/pVjlVq3Y61612v54ktiRhi3G1VgtFynchUgzLRyr/ar07p4Ya7YVK7fnolwgzxy/F4rVsAdJEV4yAyxaLVfSjga44wHSygc5pkocUuV43VViSfiiEKu2O0pX1OtDpWg345ZjKt1PToMrIABAEAQgAEAQBCAAQBAEIABAEAQgAEAQBCAAQBAEIABAEENbB9RbmZiydW6yftnS+eU91fpAjNVy0VYhH638c/9FzUE5outbfGt9xhx1QKqGJRc1Ba5an0GWjFd1QrZl388Ycy3N71nn0C3sL5GZ7Nw6tbP1Wva+CfmoqsMK83uhqlFRbSlqpThnSlW3VN+v2LZS2zr2K3+9Fi0gGuJ0qtS2rqdOnG+Vow6oFD1UVO2eq9Ztqdo8LARXQACAIAhAAIAgCEAAgCAIQACAIAhAAIAgCEAAgCCGNg07H4tN1T43PpY689XkoyIdsaO3LdoivVjMVJXolM6yJcZTv/Th5n5lmwFH6qv3veo04Uwck0rfrsWqlYNjW1erB1+uFFYlTezH1BEp2skA86TGarl4PGpZ/rxIvNP6XXOY5/bxPLO/8Iquo6fCICrfdhj++60c1wOVeG7LSLwmHSn0ius1aztX82xhr0eugAAAQRCAAABBEIAAAEEQgAAAQRCAAABBEIAAAEEQgAAAQQxtHVA2FpmyHS2qHqdWjNjHXDVEKhyrFctlnU+9ragvcv0KoJZoL8Qy9pmjbkMt3x6uHcAgNURLcEBNuwb/Zezbqb2VxljL3i5jNNWtNNpx7l0H1CvtL6CeOGeyUp9PqtZH1R657nd6tm0d6xr7WK3MResDR8sRI14fcS/2rgmU71+RPqQq8XvNZr1oydoxpEn/c7zI7ef+2bgCAgAEQQACAARBAAIABEEAAgAEQQACAARBAAIADH8a9rZt28xf//Vfmx/+8IdmdHTU/NzP/Zz5oz/6I3PFFVfM/czs7Ky5//77zY4dO0y32zWbNm0yjz32mFm9evWiDqzOFo36pE270rBFlqmpUkeKpG84dmUxqtRYx6rlZaZSrcXy+b3EO71bjdUiNRzbH2uW6mPKRNpy6mhb0U5zr3TotiWNdCFpza704o44ppWtrnVsxJGGrdKWZwtdazCdtb1S913p3Wq8NcAcdxP7CzpztOCoYpEuvbAs4f7bRoO8F5glueMy933BDtaaxZZ+XyxFO4Y9e/aYO++80+zbt888++yzJssy87GPfcxMT0/P/cx9991nnn76abNz587m5w8dOmRuvfXWxdwNAGAZWNQV0DPPPDPv66eeesq8613vMvv37ze/8Au/YI4dO2aeeOIJ841vfMPccMMNzc88+eST5n3ve18TtK677rrze/QAgOX5N6A64NRWrVrV/FsHovqqaOPGjXM/c+WVV5p169aZvXv39t1H/THd8ePH590AABc+7wBUlqW59957zfXXX2+uuuqq5nuHDx827XbbTE1NzfvZ+u8/9Zjt70qTk5Nzt7Vr1/oeEgBgOQSg+m9B3//+95tkg0Fs3bq1uZI6czt48OBA+wMAXMCLkd51113m29/+tnn++efNZZddNvf9NWvWmF6vZ44ePTrvKujIkSPNWD+dTqe5AQCWl0UFoKqqzN1332127dplnnvuOXP55ZfPG1+/fr1ptVpm9+7d5rbbbmu+d+DAAfP666+bDRs2LP7aLF58qnSV+I0NlCJZDZC3WfhvW4k0bCNWAm6ILMnYkYatVvxVY4Vj1fCyLdJmR3LvNOxRsfL0xSP/m8HZTypSfVU6tGvblak9Dbt0nIjHe/Yl3380Oyq3ne62vVZX77T0/I+27XPccqRLJ2JcjUUi5f/0D+hhva3YtyjncK2GbdR47CgTka9LtVy/8V4ZXxcE1OdM/21LsWK4dwCqP3arM9y+9a1vmZUrV879Xaf+201dF1T/e/vtt5stW7Y0iQkTExNNwKqDDxlwAADvAPT44483/370ox+d9/061frXf/3Xm/9/+OGHTRzHzRXQ2YWoAAAM9BGcy8jIiNm+fXtzAwDAhrXgAABBEIAAAEEQgAAAQRCAAADvnELUn4Q4q5dp7zPgyIOIxWr0rlYOiuxAkOmCg1Ku3+4vEjVEkaoRGpSqK1BtHhwFCZUo1KraumBKtVxY2bbX3Fzc0XVAiagHOZ7Z63FqvdL+eE7k9uLrk5kuzH7r1Lh9v6f0MWWZmGPV3qOta556eerfjkHU+sz27C/oIne0HMnF79eu14fna9b1uqvkSyBy7Fzt2HirCvvGrqYKVWqpA3I8N2dwBQQACIIABAAIggAEAAiCAAQACIIABAAIggAEAAhiaNOw01OVSfJz0wNLkWZdq/rmbp/hSJduyxxhMaT3G4k1zSNHnqMal60pXG0rxLhzSXk5xfZtI8c8VYXn8vjGmPF2zzo21T5lH2vNyP3mpX2iTkbqhDHmVD7q1Tbh+IxOpZ6ZES0Vuo70V5WaLKa419P7zVL/tgmROKRStBWRadY1lRLtaDmiSw3sQ5GjJCNRbQoifUhqXL60HPNftsV7m+clSiVae5yNKyAAQBAEIABAEAQgAEAQBCAAQBAEIABAEAQgAEAQBCAAQBBDWwfUmi5N2js34b5s+bcZcC2VXqj2BpZlx2uii8CPNxZDzvXOjdf9lqmjlicZYCl6VScktq2iAY7JoRD1OrOFvXjsaGavx6mVohDiVK7rgE72Ol61PjOndDuGYkZMVBb7vwbEmOupk+dpy1GHos5VVXMzSC2Pq3ZPbmsfinuuOiDjrYp9awL93zPLWJ9PthYe1SztGAAAQ4wABAAIggAEAAiCAAQACIIABAAIggAEAAhiaNOw2ydKk7b6pGGLdOhaLJZozxwpkLkYLzoivXiA1gcDtU0o/VPOVQq3qx1Dldj3XYr9Fq7UcLGkv8uJrj0luqwmrGM/6o5532dW6FTT6Z79mLIs8WpB0BDnuKsdQCzG1Vgy49qvfSwf08970fFLh3amhitLtG2kWoq4XpfRAHc8wFaRON/iWce+bSnes7RjAAAMMQIQACAIAhAAIAgCEAAgCAIQACAIAhAAIIihTcNOZ3OT5vk53y8Tx2q/lT3pMBIrJjfjhX08zirv1bBV6njZcm3ru6K1GYAjhTsWczzArztRn7T7uTFXZrJIiZ4RKcKZ45zwXYG71s3tx5SLsUqkWTcGSL+XKbeiDCGd0YeUzA6y8rTf6uouMk3b8frwXQ3buV/1UGO9rRHlD/qYKv9jcq04bhmOugt73rgCAgAEQQACAARBAAIABEEAAgAEQQACAARBAAIABEEAAgAEMbR1QKao+iaox6oIos5bF0vKp45tVe5/Ifabi1YNzX7bKg/fv87Bv3mBMZV45ktH2wRVm1S1RL2UY7+JGE9Tvc59EttnI03s27bEdrVS1KE4Vt43hagrq1S9jqMdg6r1kfUrzR07xj23U/erWjWc3tiv1Ymrbka2cnA8eXHuN+aiypoqx7uxfN0lHi0T5sbFoKj5k3VAC5wjroAAAEEQgAAAQRCAAABBEIAAAEEQgAAAQRCAAABBDG8adp2a6VqDv99meeUfbUvRDkClzTryQXORaq1aGzTjYgl2lWbqahEh00Fd2yb+2yplZp/HrmnLbXtde1+LE+o+HcvNq9YIlTjehkjdj7v2bRNnGrbxTsOW6cU9v1YNp8dFCr1rjsW+VbsSVyuTgdKwM7/H6iSmomw53u86Ylt1SK50dZnq7lcmosphzsYVEAAgCAIQACAIAhAAIAgCEAAgCAIQACAIAhAAIAgCEAAgiKGtAyo6sYnS8xwfZbK8MbEYjxy1DFrsnfsvl6OXS7A7DmmAh9OnS8b/3q9alv+UPqho2j4eOeoKVN1G+5h925H/1udEOqvOCbmpfH7ykchrrFaqehDHK1rVv6g5TE/peWrNqHly1M2I4XxUzZNjt6KGxVWvFmeVVy2Vq75InTOlbNui34MKcU4UuoROzv9C2yqcs0tXC44f4woIABAEAQgAEAQBCAAQBAEIABAEAQgAEAQBCAAw/GnYjz/+eHP793//9+brn/3ZnzUPPPCA+fjHP958PTs7a+6//36zY8cO0+12zaZNm8xjjz1mVq9evegDK9PYlK14USnAzbhKpS7800Gjyj5YiZTNWpyqY4r8l5T3HGuOSaZJOlLD1VL14leapKv3m8zax1on9Ry3xfjYYft6/6P/97/lfqsT08ZXtHLcOpa9e8o61l2l82a7E/ZJLjr+51MsWpmodPRaIsaTnj4Z48w+nq2wv0Vl4442KCpd3ZWarOapGKDNg3wPiuS2qnVLqdq2ONpWDNBcwvp+7Hqf9roCuuyyy8yDDz5o9u/fb1566SVzww03mJtvvtm8+uqrzfh9991nnn76abNz506zZ88ec+jQIXPrrbcu5i4AAMvEoq6Abrrppnlf/+Ef/mFzRbRv374mOD3xxBPmG9/4RhOYak8++aR53/ve14xfd9115/fIAQDL829ARVE0H7VNT0+bDRs2NFdFWZaZjRs3zv3MlVdeadatW2f27t1r3U/9Ud3x48fn3QAAF75FB6BXXnnFrFixwnQ6HfPZz37W7Nq1y7z//e83hw8fNu1220xNzf9su/77Tz1ms23bNjM5OTl3W7t2rd8jAQBc2AHoiiuuMC+//LJ54YUXzB133GE2b95sfvCDH3gfwNatW82xY8fmbgcPHvTeFwDgAl6MtL7Kec973tP8//r1682LL75ovvzlL5tPfvKTptfrmaNHj867Cjpy5IhZs2aNdX/1lVR9AwAsLwOvhl2WZfN3nDoYtVots3v3bnPbbbc1YwcOHDCvv/568zeixarTnvulVDvTsFWaoysLW608nS7NitaJI4XbnLQPlSK90nVMOs3UsSpv7pei2prWOartE/aDap3Qy+smJ7v2Yzpqn8Tyrf+R+y1nRG54qR9PktsnKlk5ah1rtR2rhpf2pZzL1LFquEi1jsRY0nWkUvcKr7HT92vfdySW/o4cS3/HuX0e89zxmhVpzb4r1NcKsd/S8W6sUq3V/YrT5fS2S9ATwVXxcka62I/L6pqfOrHgxIkTTcbbc889Z/72b/+2+fvN7bffbrZs2WJWrVplJiYmzN13390EHzLgAAADBaA333zT/Nqv/Zp54403moBz9dVXN8Hnl3/5l5vxhx9+2MRx3FwBnV2ICgDAQAGorvNRRkZGzPbt25sbAAAKa8EBAIIgAAEAgiAAAQCCIAABAIJYggzw86Ou9+lb8yPaLTTbOcaVSiyHXjnKahRVGxNP6+NtiXG19H7hWG4+6Yql9x01H2rp/XRG1L4ct7dFaMaPidYHp2bktuX0KetY0bPfb5WJoqZmx4719ZVKzKNo7+E6h5NZ+34dJR8mVjU3om1CnDlqecS4GmsU4n7b9keUiNq8WhVHXjU1zbgoGlT7VTVCzXjkX0NkVP2Rul/XMSXqsTqOyfJ4qgW+D3MFBAAIggAEAAiCAAQACIIABAAIggAEAAiCAAQACGJo07CrKOqb7hgNkPnqTG/NRHqrSE127TdSaaazOg1YpbAW4/Zc63yFzsOOxTG51lJXLS9kuq4jbTa/ZKW4z3F9TCq9WG0YO1J5O6L1gRirdcfsL69spX3bXKTXN8ckDjnSGfQm6VVeY6qNg/OcEOndrn3nY2KexvRz1xsX7RjG5KYmKjzHHPMfq5ddorf1LQVRx9uMR/YdR75p5Y52F2dwBQQACIIABAAIggAEAAiCAAQACIIABAAIggAEAAiCAAQACGJ464BiW72Dzi+P+vZw+PGYWAK/GRd1QHHPnkwfn9JtBqJTs9ax6sRJuW0l2gykF01Zx5KJFXK/JrUXHVSOeh1JzHE+MSI3zSZa9t06ls9XypZqW6H3m43Zx7MVetuiY/yWuR+g9Ufk6C6Rztp3Hos6IFctiWo5olp/uOqA1PPjfO7E85O5Xh4zfi1SnPMkHmvpOsd964BctUninHG1Y7CVEFUL7GLCFRAAIAgCEAAgCAIQACAIAhAAIAgCEAAgCAIQACCIoU3DrlNn+6XPOpcWL+25ipUYayT2eFyK7gZV3NHHNGpPL44mxvzbMYzZD6oQ93l6W/tTn4/q30v0cvT2NNOeaEFQ607an58y9U9RrcTdlo5XgHreVZr16W0rr2OqBkirdb0+ilGVhm0fS+yVBM5U67q1ity255dCX7nevSL/1GR5jou05ThztXzxPxeN68R4B+IKCAAQBAEIABAEAQgAEAQBCAAQBAEIABAEAQgAEMTwpmGnUd/U28i1Omshci8dadiRSBeNRBpwNaqncZCVnCuVXiyOSaWv1roiJdq1UrBKUVUr6/Ym9DH1Ju1jZUvnoMpVe2O/+W3GE7F6sc50NyZVadj+ObXyHM8dK3SLNOE4s2+bnnSlwatxZ2K5fcvUP225WqI0bHWOqzTrZrxnv+Oi7boeGGCZ9CHFFRAAIAgCEAAgCAIQACAIAhAAIAgCEAAgCAIQACAIAhAAIIjhrQPqRKbwaMegCoXK9gD1OLF921wscV8r1P06SiQiMV6qJf0dz6w65nzUeNfcqLF8XD9YOR47JkoNR/6/glWpKBhp62KSuGUfT8TjcVXNlD37E1919QMqY7+iqGKk8m6DomuE9EmuXney9svx2olFC4hmPPer03LWJon5LxzvT6qFhzzHXQ9FjLvm2Ha/rvq6M7gCAgAEQQACAARBAAIABEEAAgAEQQACAARBAAIABDG0adhNCnaftETXMuoqVVGmijpSDvulhJ/Rm9T7zcbVMclN5bhvOnStbKsxnbdZdESLApGuW3UcacujuXcqaZn7/S4VOdK7U5FK3RnRa++309wrDTsv9GOZSe19IHqx7hFRifzYqhDPqyOvNvfOg1/IuJ9Itk1wpJW7yj0sSkfrlUo8PYV4TQ6Shq3S0Zv9qkHPdiULbTfCFRAAIAgCEAAgCAIQACAIAhAAIAgCEAAgCAIQACAIAhAAIIihrQPqTkYm6Sy+DigWpRnuVg72oTIV7QvG9G6LEf9jku0YxLNXpY5aHlGvU4w6Jlm0IYjEWKulH2wqxktHDVeRqCIJu8hRJJGIdgxJXHrvuxCPJyv0Y8kz+xNfZf49CkrVXsJRWKZq7FytQWQLg8q/vkX9eq3aPCxs537tGFT9XdHn/W6h+1ZPj7MOaIB6Quv4Ai9tuAICAARBAAIABEEAAgAEQQACAARBAAIABEEAAgC889KwH3zwQbN161Zzzz33mEceeaT53uzsrLn//vvNjh07TLfbNZs2bTKPPfaYWb169aL23Zs0JhlZfNpynEVey7O7qKXQ81Hj3b7Ame2p0rBb/i0VqnH7ZLTGM+906TQVY4605UFaFGQiXbryy6g9fUyJfb9p4rlmf5OGbX88WabTsIuuGHelYatl8sX8V6WexDK3v+7KlqP1gdhWll24WpkMkF7s2yFCtkxwpFqX7cFSvO0bOsblPDneR2LP+XXftfbiiy+ar33ta+bqq6+e9/377rvPPP3002bnzp1mz5495tChQ+bWW2/1vRsAwAXKKwCdPHnSfOpTnzJf//rXzUUXXTT3/WPHjpknnnjCfOlLXzI33HCDWb9+vXnyySfNP/zDP5h9+/adz+MGACzHAHTnnXeaT3ziE2bjxo3zvr9//36TZdm871955ZVm3bp1Zu/evX33VX9Md/z48Xk3AMCFb9GfKtZ/2/ne977XfAT3docPHzbtdttMTU3N+3799596rJ9t27aZ3/u931vsYQAAltMV0MGDB5uEgz//8z83IyNigbNFqJMY6o/uztzq+wAAXPgWFYDqj9jefPNN88EPftCkadrc6kSDRx99tPn/+kqn1+uZo0ePztvuyJEjZs2aNX332el0zMTExLwbAODCt6iP4G688UbzyiuvzPvepz/96ebvPL/zO79j1q5da1qtltm9e7e57bbbmvEDBw6Y119/3WzYsGFRB5avKE05Ui5q1d1mXGXGutJxI7/0SpVm3WzbcuVBelIptY7U12REpGG3db566pmarNKsT4/7rzzdEunfauXpUqRDu1a0dj0ezX68nc4Ae+34rQpeKwuxorUYa8bFuahWkq9FopxgEOq9wLWqfj4eee3XlYZdivHKuWq4Whp8id73HMdURX7z4BWAVq5caa666qp53xsfHzcXX3zx3Pdvv/12s2XLFrNq1armaubuu+9ugs911123mLsCAFzgzns/oIcfftjEcdxcAZ1diAoAwHkNQM8999y8r+vkhO3btzc3AABsWAsOABAEAQgAEAQBCAAQBAEIAHBhZMGdL8V4aapRj/qZaoAl1lVdjRiLRF1MLVbbOo5J1aHE4n5jV81N4l9z41vr49pvIh6rq/WBmifV+qBX6IKF0lboUM+xo5eGOqaWOCdGWroOa7zTs47ljsfTze3js7P2gpxMtYBwLenvqo0RdXRyWX/Plgmnd+zf+qASNXauli+q9YRxPB49F+IBiXPYua2LrQ5I1SydhSsgAEAQBCAAQBAEIABAEAQgAEAQBCAAQBAEIABAEEObhh2N5SYa7ZPT6Mjuqwr/mKrSqWORVhg70otVurRKh3a1PtDtC/RElQNkXsYiq1OlJrcc89RJ7TmsseOJL0UOqzomVyq1b5r1Ut5vrtLKI//nPU/t+dJZqp+7KrUfU+VITVatHlQGsSu929VyQR6TeL2XHfuOI1fricz/mIxny4XI9WIfoHTF2o5hgSnyXAEBAIIgAAEAgiAAAQCCIAABAIIgAAEAgiAAAQCCIAABAIIY2jqg9khmktFzE/3LUieY55m9OKBybBuJ2hnftgi1VNRQdFq6MEAtza/aF7haH2Ri2X5XiwJfrmMaT3tedT61bp761Salus1DKo45dRSaxJ6FKLmjwGU6a5slqeFS579qVdLU5KiWCgP0TVCbOmrdjHi9D9KBQB2Tsw3BIC0kFNHmIdKnuC7aGWii3LgCAgAEQQACAARBAAIABEEAAgAEQQACAARBAAIABDG8aditwiR9UpALsRR9rRJZg6VjW5Wuq1K0XS0VVPpxy7FtO7HnULbjwjvlWaUXt8R91iqRtlmKMXWfzbYD5Ki6jtmXSrV2pVmXlf18y8WYSpF3jbu27eX28WKAViYyNdmR1V+2xLatymusOSRR4RBl+lyT6dQqJd2ReS+edhM5ykRky4WB0sqXNtVa4QoIABAEAQgAEAQBCAAQBAEIABAEAQgAEAQBCAAQxPCmYaeFSfqsVJwXOmWwTO2pjIUjU1etBhyLFOJB0rCXKl26HdtX0a6NJJlX+rArhVilYasx1yrcrhRu1+P1XXlapYaXjm1zkfY/k7e8VyOXadgizdo1LleSLyLvVN6q5UhXTyK/bR1p2FVin/9I3Gezbdt+v5E4pkqsSt1QqdaV3jRS+y4DrMB9HnAFBAAIggAEAAiCAAQACIIABAAIggAEAAiCAAQACIIABAAIYmjrgFpJbtL03LqEJPaPmYWoEXLVmiSiRihyLGcu2zE46ltiV3GAbbsBllhPHTU1qajnycvEu92CbOXgaH2g9+tXt+Qad22biblQtT6uliOKqlc7Pa5ajojaF1FT42xf4OoyoHqoiFqfuKUL+9RMVK7fvcXjiWTdn6tdjKiXMg6iFitS5UWu00nd8RJ3auAKCAAQBAEIABAEAQgAEAQBCAAQBAEIABAEAQgAEMTQpmG3k8KkfdoNVI40U1d7A6Xl2frAN1V6UKVnOnQtFmnNrhRuNZ7GxZLs193KIfVqH6FaJtSKAVK4VdsEta0rrX9EnIuuY1oqRSQeax57tyhQKc+RY7dq28rxzhclKl1dbVh6H1PZTbzTsOXT7rrMEIccudpwDPjWxxUQACAIAhAAIAgCEAAgCAIQACAIAhAAIAgCEAAgCAIQACCIoa0Dqlsj9GuPMEidg6u+opPksi5paep1XMu3+z1eV+uDpfrNQ9YIOWokVA1Rr0yXZP5d86u2dbVN8D1XXfVSqtbNUUkiz7e8ELVhjjYopaqbKfXjUc9BJNpHGFfLEdWjwPUC8LxfdZdOsWM8WoIx1/36v+0tCFdAAIAgCEAAgCAIQACAIAhAAIAgCEAAgCCGLguuqk5nmOSnel6ZRSozyZUFl4gsOLWKrYvOpIq8M4TUyuCVyuJpVs+1p7e4Vhz3VTmy4CqRBZeVetssr7wyv1yZbGolbee2YlydE6Vr1fABMjKLrPQb69lfG7Uy918NuyrEeFZ4rVjd3K/Yb+U4pigV59sA7wVK2dVvx9GMyHEsB7jMUAl/ud9q2OXs7Lz3c+v+K9dP/IT953/+p1m7dm3owwAADOjgwYPmsssue+cEoLIszaFDh8zKlStNFEXm+PHjTUCqH8jExETowxtazNPCME8LwzwtDPPUXx1WTpw4YS699FJZQzZ0H8HVB9svYtZPLk+wG/O0MMzTwjBPC8M8nWtyctK4kIQAAAiCAAQACGLoA1Cn0zFf+MIXmn9hxzwtDPO0MMzTwjBPgxm6JAQAwPIw9FdAAIALEwEIABAEAQgAEAQBCAAQxNAHoO3bt5uf+ZmfMSMjI+baa681//iP/2iWs+eff97cdNNNTYVxvVLEN7/5zXnjdU7JAw88YN797neb0dFRs3HjRvOv//qvZjnZtm2b+fCHP9yspvGud73L3HLLLebAgQPzfmZ2dtbceeed5uKLLzYrVqwwt912mzly5IhZTh5//HFz9dVXzxVRbtiwwfzN3/zN3Dhz1N+DDz7YvPbuvffeue8xVxdgAPrLv/xLs2XLlibN8Xvf+5655pprzKZNm8ybb75plqvp6elmHurA3M8Xv/hF8+ijj5qvfvWr5oUXXjDj4+PNnNUvkOViz549zZvBvn37zLPPPmuyLDMf+9jHmrk747777jNPP/202blzZ/Pz9fJPt956q1lO6hVH6jfT/fv3m5deesnccMMN5uabbzavvvpqM84cnevFF180X/va15rAfTbmylM1xD7ykY9Ud95559zXRVFUl156abVt27agxzUs6qdv165dc1+XZVmtWbOmeuihh+a+d/To0arT6VR/8Rd/US1Xb775ZjNXe/bsmZuTVqtV7dy5c+5n/uVf/qX5mb1791bL2UUXXVT9yZ/8CXPUx4kTJ6r3vve91bPPPlv94i/+YnXPPfc032eu/A3tFVCv12t+M6s/Qjp7nbj667179wY9tmH12muvmcOHD8+bs3o9pvqjy+U8Z8eOHWv+XbVqVfNvfV7VV0Vnz9OVV15p1q1bt2znqSgKs2PHjuYqsf4ojjk6V31V/YlPfGLenNSYK39DtxjpGW+99Vbzoli9evW879df//CHPwx2XMOsDj61fnN2Zmy5qVdXrz+rv/76681VV13VfK+ei3a7baampsxyn6dXXnmlCTj1R7T13y527dpl3v/+95uXX36ZOTpLHZzrPwPUH8G9HefTBRiAgPP1W+v3v/998/d///ehD2UoXXHFFU2wqa8S/+qv/sps3ry5+RsG/lfdauGee+5p/p5YJ0Ph/Bnaj+AuueQSkyTJOZkk9ddr1qwJdlzD7My8MGen3XXXXebb3/62+e53vzuvxUc9F/VHvEePHjXLfZ7q39zf8573mPXr1zfZg3WCy5e//GXm6G0fsdWJTx/84AdNmqbNrQ7SdbJP/f/1lQ5zdYEFoPqFUb8odu/ePe/jlPrr+iMDnOvyyy9vTviz56xumFVnwy2nOavzM+rgU3+c9J3vfKeZl7PV51Wr1Zo3T3Wa9uuvv76s5qmf+jXW7XaZo7PceOONzUeV9ZXimduHPvQh86lPfWru/5krT9UQ27FjR5PB9dRTT1U/+MEPqs985jPV1NRUdfjw4Wq5qjNx/umf/qm51U/fl770peb//+M//qMZf/DBB5s5+ta3vlX98z//c3XzzTdXl19+eTUzM1MtF3fccUc1OTlZPffcc9Ubb7wxdzt16tTcz3z2s5+t1q1bV33nO9+pXnrppWrDhg3NbTn53Oc+12QGvvbaa825Un8dRVH1d3/3d804c2R3dhZcjbnyM9QBqPaVr3yleWLb7XaTlr1v375qOfvud7/bBJ633zZv3jyXiv35z3++Wr16dRO8b7zxxurAgQPVctJvfurbk08+OfczdUD+rd/6rSbteGxsrPqVX/mVJkgtJ7/xG79R/fRP/3Tz2vqpn/qp5lw5E3xqzNHCAxBz5Yd2DACAIIb2b0AAgAsbAQgAEAQBCAAQBAEIABAEAQgAEAQBCAAQBAEIABAEAQgAEAQBCAAQBAEIABAEAQgAEAQBCABgQvj/6utcyDqGPOsAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1. 0.]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0. 1.]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0. 1.]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for imgs,labels in test_batches: # visualization of test images\n",
    "    for i in range(10):\n",
    "        print(labels[i])\n",
    "        img = cv2.resize(imgs[i],(50,50))\n",
    "        plt.imshow(img)\n",
    "        plt.show()\n",
    "    break"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model Defination"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Library/Frameworks/Python.framework/Versions/3.12/lib/python3.12/site-packages/keras/src/layers/convolutional/base_conv.py:107: UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.\n",
      "  super().__init__(activity_regularizer=activity_regularizer, **kwargs)\n"
     ]
    }
   ],
   "source": [
    "model = keras.Sequential([\n",
    "    layers.Conv2D(32, kernel_size=(3, 3), activation='relu', input_shape=(24,24,1)),\n",
    "    \n",
    "    # A convolutional layer receives input,transforms the input in some way,and then outputs the transformed input to the next layer.\n",
    "    # With each convolutional layer, we need to specify the number of filters the layer should have. These filters are actually\n",
    "    # what detect the patterns like edges,objects,colors,shapes,curves etc.\n",
    "    # In even deeper layers, the filters are able to detect even more sophisticated objects like full dogs, cats, birds.\n",
    "    # A filter can technically just be thought of as a relatively small matrix ( tensor), for which, we decide the number of \n",
    "    # rows and columns this matrix has, and the values within this matrix are initialized with random numbers. \n",
    "    # we take the dot product of the filter with the first 3 x 3 block of pixels, and then that result is stored in the output channel.\n",
    "    # Then, the filter slides to the next 3 x 3 block, computes the dot product, and stores the value as the next pixel in the output channel.\n",
    "    \n",
    "    # In an artificial neural network, an activation function is a function that maps a node's inputs to its corresponding output. \n",
    "    # as we know that every connection in network has some wights ,so we calculate the sum of products of input and weights and give\n",
    "    # it to the activation function\n",
    "    # relu(x) = max(0,X) so more positive the neuron is more activated it is\n",
    "    \n",
    "    layers.MaxPooling2D(pool_size=(1,1)),\n",
    "    # When added to a model, max pooling reduces the dimensionality of images by reducing the \n",
    "    # number of pixels in the output from the previous convolutional layer\n",
    "    # The first parameter we’re specifying is the pool_size. This is the size of what we were calling a filter before, and in our example, we used a 1 x 1 filter. \n",
    "    \n",
    "    layers.Conv2D(32,(3,3),activation='relu'), \n",
    "    # another 2-dimentional convolutional layer because we are working with 2d images\n",
    "    # 32 convolution filters used each of size 3x3\n",
    "\n",
    "    layers.MaxPooling2D(pool_size=(1,1)),\n",
    "    \n",
    "    layers.Conv2D(64, (3, 3), activation='relu'),\n",
    "    # third convolutional layer\n",
    "    # 64 convolution filters used each of size 3x3\n",
    "    \n",
    "    layers.MaxPooling2D(pool_size=(1,1)),\n",
    "    # choose the best features via pooling\n",
    "    \n",
    "    # randomly turn neurons on and off to improve convergence\n",
    "    layers.Dropout(0.25),\n",
    "\n",
    "    # flatten since too many dimensions, we only want a classification output\n",
    "    layers.Flatten(),\n",
    "\n",
    "    # fully connected to get all relevant data\n",
    "    layers.Dense(128, activation='relu'),\n",
    "\n",
    "    # one more dropout for convergence' sake \n",
    "    layers.Dropout(0.5),\n",
    "\n",
    "    # output a softmax to squash the matrix into output probabilities\n",
    "    layers.Dense(2, activation='softmax')\n",
    "    #we have two output nodes which denotes either open or closed\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">Model: \"sequential\"</span>\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1mModel: \"sequential\"\u001b[0m\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
       "┃<span style=\"font-weight: bold\"> Layer (type)                    </span>┃<span style=\"font-weight: bold\"> Output Shape           </span>┃<span style=\"font-weight: bold\">       Param # </span>┃\n",
       "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
       "│ conv2d (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Conv2D</span>)                 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">22</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">22</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">32</span>)     │           <span style=\"color: #00af00; text-decoration-color: #00af00\">320</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ max_pooling2d (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">MaxPooling2D</span>)    │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">22</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">22</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">32</span>)     │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ conv2d_1 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Conv2D</span>)               │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">20</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">20</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">32</span>)     │         <span style=\"color: #00af00; text-decoration-color: #00af00\">9,248</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ max_pooling2d_1 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">MaxPooling2D</span>)  │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">20</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">20</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">32</span>)     │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ conv2d_2 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Conv2D</span>)               │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">18</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">18</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">64</span>)     │        <span style=\"color: #00af00; text-decoration-color: #00af00\">18,496</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ max_pooling2d_2 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">MaxPooling2D</span>)  │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">18</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">18</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">64</span>)     │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dropout (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dropout</span>)               │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">18</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">18</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">64</span>)     │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ flatten (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Flatten</span>)               │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">20736</span>)          │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dense (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                   │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)            │     <span style=\"color: #00af00; text-decoration-color: #00af00\">2,654,336</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dropout_1 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dropout</span>)             │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)            │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dense_1 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">2</span>)              │           <span style=\"color: #00af00; text-decoration-color: #00af00\">258</span> │\n",
       "└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
       "</pre>\n"
      ],
      "text/plain": [
       "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
       "┃\u001b[1m \u001b[0m\u001b[1mLayer (type)                   \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape          \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m      Param #\u001b[0m\u001b[1m \u001b[0m┃\n",
       "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
       "│ conv2d (\u001b[38;5;33mConv2D\u001b[0m)                 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m22\u001b[0m, \u001b[38;5;34m22\u001b[0m, \u001b[38;5;34m32\u001b[0m)     │           \u001b[38;5;34m320\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ max_pooling2d (\u001b[38;5;33mMaxPooling2D\u001b[0m)    │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m22\u001b[0m, \u001b[38;5;34m22\u001b[0m, \u001b[38;5;34m32\u001b[0m)     │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ conv2d_1 (\u001b[38;5;33mConv2D\u001b[0m)               │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m20\u001b[0m, \u001b[38;5;34m20\u001b[0m, \u001b[38;5;34m32\u001b[0m)     │         \u001b[38;5;34m9,248\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ max_pooling2d_1 (\u001b[38;5;33mMaxPooling2D\u001b[0m)  │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m20\u001b[0m, \u001b[38;5;34m20\u001b[0m, \u001b[38;5;34m32\u001b[0m)     │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ conv2d_2 (\u001b[38;5;33mConv2D\u001b[0m)               │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m18\u001b[0m, \u001b[38;5;34m18\u001b[0m, \u001b[38;5;34m64\u001b[0m)     │        \u001b[38;5;34m18,496\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ max_pooling2d_2 (\u001b[38;5;33mMaxPooling2D\u001b[0m)  │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m18\u001b[0m, \u001b[38;5;34m18\u001b[0m, \u001b[38;5;34m64\u001b[0m)     │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dropout (\u001b[38;5;33mDropout\u001b[0m)               │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m18\u001b[0m, \u001b[38;5;34m18\u001b[0m, \u001b[38;5;34m64\u001b[0m)     │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ flatten (\u001b[38;5;33mFlatten\u001b[0m)               │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m20736\u001b[0m)          │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dense (\u001b[38;5;33mDense\u001b[0m)                   │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m)            │     \u001b[38;5;34m2,654,336\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dropout_1 (\u001b[38;5;33mDropout\u001b[0m)             │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m)            │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dense_1 (\u001b[38;5;33mDense\u001b[0m)                 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m2\u001b[0m)              │           \u001b[38;5;34m258\u001b[0m │\n",
       "└─────────────────────────────────┴────────────────────────┴───────────────┘\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Total params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">2,682,658</span> (10.23 MB)\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1m Total params: \u001b[0m\u001b[38;5;34m2,682,658\u001b[0m (10.23 MB)\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">2,682,658</span> (10.23 MB)\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m2,682,658\u001b[0m (10.23 MB)\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Non-trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> (0.00 B)\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model.summary() # summary of our model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## training our model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(\n",
    "    optimizer='adam',\n",
    "    loss='categorical_crossentropy', \n",
    "    metrics=['accuracy']\n",
    ")\n",
    "# The Adam optimization algorithm is an extension to stochastic gradient descent\n",
    "# loss function is for calculating loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/15\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Library/Frameworks/Python.framework/Versions/3.12/lib/python3.12/site-packages/keras/src/trainers/data_adapters/py_dataset_adapter.py:121: UserWarning: Your `PyDataset` class should call `super().__init__(**kwargs)` in its constructor. `**kwargs` can include `workers`, `use_multiprocessing`, `max_queue_size`. Do not pass these arguments to `fit()`, as they will be ignored.\n",
      "  self._warn_if_super_not_called()\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "30/30 - 1s - 47ms/step - accuracy: 0.7867 - loss: 6.5671 - val_accuracy: 1.0000 - val_loss: 0.0043\n",
      "Epoch 2/15\n",
      "30/30 - 0s - 16ms/step - accuracy: 1.0000 - loss: 0.0020 - val_accuracy: 1.0000 - val_loss: 0.0000e+00\n",
      "Epoch 3/15\n",
      "30/30 - 1s - 17ms/step - accuracy: 0.9967 - loss: 0.0064 - val_accuracy: 1.0000 - val_loss: 2.3304e-05\n",
      "Epoch 4/15\n",
      "30/30 - 1s - 18ms/step - accuracy: 0.9900 - loss: 0.0297 - val_accuracy: 1.0000 - val_loss: 5.9605e-09\n",
      "Epoch 5/15\n",
      "30/30 - 1s - 17ms/step - accuracy: 0.9967 - loss: 0.0263 - val_accuracy: 1.0000 - val_loss: 0.0000e+00\n",
      "Epoch 6/15\n",
      "30/30 - 1s - 18ms/step - accuracy: 1.0000 - loss: 8.3854e-06 - val_accuracy: 1.0000 - val_loss: 0.0000e+00\n",
      "Epoch 7/15\n",
      "30/30 - 0s - 16ms/step - accuracy: 1.0000 - loss: 0.0012 - val_accuracy: 1.0000 - val_loss: 0.0000e+00\n",
      "Epoch 8/15\n",
      "30/30 - 0s - 16ms/step - accuracy: 1.0000 - loss: 0.0026 - val_accuracy: 1.0000 - val_loss: 0.0000e+00\n",
      "Epoch 9/15\n",
      "30/30 - 0s - 16ms/step - accuracy: 0.9833 - loss: 0.0681 - val_accuracy: 1.0000 - val_loss: 9.5367e-08\n",
      "Epoch 10/15\n",
      "30/30 - 0s - 16ms/step - accuracy: 0.9767 - loss: 0.1920 - val_accuracy: 1.0000 - val_loss: 0.0000e+00\n",
      "Epoch 11/15\n",
      "30/30 - 0s - 16ms/step - accuracy: 1.0000 - loss: 2.4864e-04 - val_accuracy: 1.0000 - val_loss: 0.0000e+00\n",
      "Epoch 12/15\n",
      "30/30 - 1s - 25ms/step - accuracy: 1.0000 - loss: 4.5236e-04 - val_accuracy: 1.0000 - val_loss: 0.0000e+00\n",
      "Epoch 13/15\n",
      "30/30 - 1s - 25ms/step - accuracy: 1.0000 - loss: 1.4917e-04 - val_accuracy: 1.0000 - val_loss: 0.0000e+00\n",
      "Epoch 14/15\n",
      "30/30 - 1s - 19ms/step - accuracy: 1.0000 - loss: 3.8216e-04 - val_accuracy: 1.0000 - val_loss: 0.0000e+00\n",
      "Epoch 15/15\n",
      "30/30 - 0s - 17ms/step - accuracy: 1.0000 - loss: 8.1794e-05 - val_accuracy: 1.0000 - val_loss: 0.0000e+00\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.src.callbacks.history.History at 0x15b31af00>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(\n",
    "    train_batches, \n",
    "    validation_data=valid_batches,\n",
    "    epochs=15,\n",
    "    steps_per_epoch=30,\n",
    "    validation_steps=10,\n",
    "    verbose=2\n",
    ")\n",
    "# our model will train on train_batches and will validate on valid_batches \n",
    "# we have make 15 epochs ,you can give any number\n",
    "# verbose is 2, which just specifies the verbosity of the log output printed to the console during training. \n",
    "# we have 300 training images and 10 per batch so steps_per_epoch is 30\n",
    "# we have 20 validation images and 2 per batch so validation_steps is 10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:You are saving your model as an HDF5 file via `model.save()` or `keras.saving.save_model(model)`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')` or `keras.saving.save_model(model, 'my_model.keras')`. \n"
     ]
    }
   ],
   "source": [
    "model.save('./models/myModel.h5', overwrite=True) \n",
    "# saving our model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Library/Frameworks/Python.framework/Versions/3.12/lib/python3.12/site-packages/keras/src/trainers/data_adapters/py_dataset_adapter.py:121: UserWarning: Your `PyDataset` class should call `super().__init__(**kwargs)` in its constructor. `**kwargs` can include `workers`, `use_multiprocessing`, `max_queue_size`. Do not pass these arguments to `fit()`, as they will be ignored.\n",
      "  self._warn_if_super_not_called()\n"
     ]
    }
   ],
   "source": [
    "predictions = model.predict(\n",
    "    x=test_batches, \n",
    "    steps=4, \n",
    "    verbose=0\n",
    ")\n",
    "# making prediction about test_batches"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## testing our model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import confusion_matrix # for plotting confusion matrix\n",
    "import itertools"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(40, 2)\n",
      "<class 'numpy.ndarray'>\n",
      "<class 'numpy.ndarray'>\n",
      "[1.000000e+00 7.592189e-20]\n",
      "[1. 0.]\n"
     ]
    }
   ],
   "source": [
    "print(predictions.shape) # prediction contain probability of classes for every image\n",
    "print(type(predictions)) # it is a numpy array\n",
    "print(type(predictions[0])) # it contains numpy array of probabilites of each class\n",
    "print(predictions[0]) # first element is probability that image is from class closed and second is probability for open class\n",
    "for i in predictions:\n",
    "    i[0]=round(i[0]) # rounding the probabilities to convert them into 0 and 1\n",
    "    i[1]=round(i[1])\n",
    "print(predictions[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "cm = confusion_matrix(y_true=test_label[:,0], y_pred=predictions[:,0]) # making confusion matrix object"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_confusion_matrix(cm, classes,normalize=False,title='Confusion matrix',cmap=plt.cm.Blues):\n",
    "    # This function prints and plots the confusion matrix.\n",
    "    # Normalization can be applied by setting `normalize=True`.\n",
    "    plt.imshow(cm, interpolation='nearest', cmap=cmap)\n",
    "    plt.title(title)\n",
    "    plt.colorbar()\n",
    "    tick_marks = np.arange(len(classes))\n",
    "    plt.xticks(tick_marks, classes, rotation=45)\n",
    "    plt.yticks(tick_marks, classes)\n",
    "\n",
    "    if normalize:\n",
    "        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "        print(\"Normalized confusion matrix\")\n",
    "    else:\n",
    "        print('Confusion matrix, without normalization')\n",
    "\n",
    "    print(cm)\n",
    "\n",
    "    thresh = cm.max() / 2.\n",
    "    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
    "        plt.text(j, i, cm[i, j],\n",
    "                 horizontalalignment=\"center\",\n",
    "                 color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "\n",
    "    plt.tight_layout()\n",
    "    plt.ylabel('True label')\n",
    "    plt.xlabel('Predicted label')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[10 10]\n",
      " [10 10]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cm_plot_labels = ['closed','open']\n",
    "plot_confusion_matrix(cm=cm, classes=cm_plot_labels, title='Confusion Matrix') # plotting confusion matrix"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "as we can see our model is not predicting well<br>\n",
    "it has predicted 10 open eyes as closed eyes and 9 closed eyes as open eyes<br>\n",
    "the reason can be the size of our dataset which is very small and all the images are taken in same intensity of light<br>\n",
    "if images are taken in large amount and in different intesities it can predict well."
   ]
  }
 ],
 "metadata": {
  "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": "3.12.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}