final_results.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "1d84ce38",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import os"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "8769d0fb",
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.read_csv(\"results.csv\")\n",
    "data = data.round(4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "26f64ea4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>model</th>\n",
       "      <th>mAP</th>\n",
       "      <th>mAP_COCO</th>\n",
       "      <th>inference</th>\n",
       "      <th>FPS</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>yolov3</td>\n",
       "      <td>0.9949</td>\n",
       "      <td>0.8862</td>\n",
       "      <td>0.0312</td>\n",
       "      <td>32.0075</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>yolov3-tiny</td>\n",
       "      <td>0.9863</td>\n",
       "      <td>0.7198</td>\n",
       "      <td>0.0156</td>\n",
       "      <td>64.0144</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>yolov5l</td>\n",
       "      <td>0.9950</td>\n",
       "      <td>0.8757</td>\n",
       "      <td>0.0312</td>\n",
       "      <td>32.0077</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>yolov5m</td>\n",
       "      <td>0.9949</td>\n",
       "      <td>0.8553</td>\n",
       "      <td>0.0156</td>\n",
       "      <td>64.0172</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>yolov5n</td>\n",
       "      <td>0.9907</td>\n",
       "      <td>0.7586</td>\n",
       "      <td>0.0156</td>\n",
       "      <td>64.0166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>yolov5s</td>\n",
       "      <td>0.9947</td>\n",
       "      <td>0.8278</td>\n",
       "      <td>0.0156</td>\n",
       "      <td>64.0139</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>yolov5x</td>\n",
       "      <td>0.9950</td>\n",
       "      <td>0.8824</td>\n",
       "      <td>0.0469</td>\n",
       "      <td>21.3386</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         model     mAP  mAP_COCO  inference      FPS\n",
       "0       yolov3  0.9949    0.8862     0.0312  32.0075\n",
       "1  yolov3-tiny  0.9863    0.7198     0.0156  64.0144\n",
       "2      yolov5l  0.9950    0.8757     0.0312  32.0077\n",
       "3      yolov5m  0.9949    0.8553     0.0156  64.0172\n",
       "4      yolov5n  0.9907    0.7586     0.0156  64.0166\n",
       "5      yolov5s  0.9947    0.8278     0.0156  64.0139\n",
       "6      yolov5x  0.9950    0.8824     0.0469  21.3386"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "e47eeccc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAakAAAEQCAYAAAAK6YvmAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAABFMElEQVR4nO3de1yVVb748c+CzVUxE1AGTYnQkkwMyUKtvJBmpYEzahz0MMboeNRGp8v85pyyi03NJRszZ7TUMbKwc2oS1JmmwducI9M0E5oMBGV5I0RJYwKV24b9/f3BZgeyEVC2bPD7fr32i/2s51nr+e5F8XWt59nPMiKCUkop5Y48OjsApZRSqiWapJRSSrktTVJKKaXcliYppZRSbkuTlFJKKbelSUoppZTbsnR2AB0pKChIwsLCOjsMpZRS7bRv377TIhJ8fnm3SlJhYWFkZ2d3dhhKKaXayRhzzFm5TvcppZRyW5qklMtU19Z1dghKqS5Ok5RyiaOnz3Hz8h0c+/pcZ4eilOrCNEkpl1iR+RmV1jpW/Pmzzg5FKdWFaZJSHe7o6XPsyC9BBDLzS3Q0pZS6aJqkVIdbkfkZtbb6p+vX1tl0NKWUumiapBqRWltnh9DlNYyi6uxJqk5HU0qpS6BJyq7ubA3Fz/2durM1nR1Kl9Z4FNVAR1NKqYulScquMv9rpLKWqvzSzg6lyzp/FNVAR1NKqYulScquIrsEgHP7TnZyJF2Xs1FUAx1NKaUuhiYpwFZhpeb4WQBqis5iq6zt5Ii6npZGUQ10NKWUuhjd6tl9bfWvLQepyP0aGv6e2gTjaZC6+p8nfv4P8DD1+wz43xTE1dMHd1q8XcGFRlENGkZTq/8t+jJFpZTq6q7IkVTP2wfg2cMLsdYhVbVITR1SU39nn9TY6rerahFrHZ49veh5e/9Ojti9tTaKaqCjKaVUe12RScor2J9+S6PpMSoE4+W8C4yXBz1GhdBvSTRewf6XOcKupS2jqAZ6bUop1R5X5HQfgLF4cPX9EXj29ObM/xYhNd8+DNV4exBw5zX0mjiwEyPsGipr6vhT7km8PA0+lvMTfi3n/ydmswnv5Z3kVzV1+Hl7XrY4lVJd0xWbpBpUHyv/NkF5ALb6Kb+awvJOjaur8PP25G//NYFqa9MvQpdV/4t5e2awfvw7XOVzdZN9Pl4emqCUUm1yRScpW00d1Ye/Aeqn93xv6ENVQSlSa6Pq0DeItQ7jpX9MW9M3wLdZ2d8PvsdZ6xk+O/Mh3/3OdzshKqVUd3BFXpNqUH3wX1AreAR4Ezx/OIFJQwn+4XA8ArygVqg6+K/ODrHLyvgio8lPpZS6GC4dSRljfgz8gPqbvXOBucDrwPX2Q3oD34jICCd1jwJngDqgVkRiOjq+2m+q8RsexNXfHYyHT31XeF8TQMgjMfxry+fU/qu6o095RSirLiP/63wAPvn6E8pryunl3auTo1JKdUUuS1LGmP7Aj4BIEak0xrwNPCAisxod8yJQdoFmxovIaVfFGDC2P9D89nIPXwuB/zbUVaftdp7+4Gl2HNuB2L94Vmerw+JhwWqzYvGwcNc7d+HpUT9tajBMGjSJp0Y/1ZkhK6W6CFdfk7IAfsYYK+APFDfsMMYYYCYwwcUxKBf7/o3fZ1/JPo6fPY7VZm2yr7K20vHey8OL/j37k3xj8uUOUSnVRbnsmpSIHAdWAIXACaBMRDIbHXI7UCIin7fUBJBpjNlnjJnvqjjVpQu7Kox3p73Ldwd/F1/P5jdRAPh6+vK9Id9jy7QthF0VdnkDVEp1WS5LUsaYq4H7gWuBUKCHMWZ2o0MSgbcu0MQYEYkGpgCLjDF3tHCe+caYbGNM9qlTpzooetVe3p7ePH7b46TclIK/pemXn/0t/qTclMJ/3fpfeHl6dVKESqmuyJV398UBR0TklIhYgS3AaABjjAWYDvxPS5VFpNj+8ysgHRjVwnHrRCRGRGKCg4M7+COo9sr5KoeK2goAPE39daiK2gpyTuV0ZlhKqS7KlUmqELjNGONvv/40ESiw74sDPhWRImcVjTE9jDEBDe+BSUCeC2NVHaCytpJ/nPwHAD6ePsQNjMPH0weAf5z4B1W1VZ0ZnlKqC3LlNam/A78H9lN/+7kHsM6++wHOm+ozxoQaY96zb/YDsowxOcA/gD+KyPuuilV1jA+Of0CNrYYgvyBS705lxbgVpN6dSpBfEDW2Gv5a/NfODlEp1cUYkbY9GLQriImJkezs7M4O44r1Rv4b5JzK4ZnRz9DDq4ej/GzNWZ7+29NEBUcxJ3JOJ0aolHJXxph9zr4Pq0lKKaVUp2spSV3Rj0VSSinl3jRJKaWUcluapJRSSrktTVJKKaXcliYppZRSbkuTlFJKKbelSUoppZTb0iSllFLKbWmSUkop5bY0SSmllHJbmqSUUkq5LU1SSiml3JYmKaWUUpdEampc1rYmKaWUUhettrSUg2Nvp7a01CXta5JSSil10c7s3ImtvJwzu3a5pH1NUkoppS5a2Zb0Jj87miYppZRSF6WurIyqTz4BoCovj7ry8g4/h6XDW1RKKdUtnVj2JOV//rNjW2prwWIBqxW8vPh83HiM5du00uvuu/nO8mcu6Zw6klJKKdUmfR6ciyWwD1JZia28HKmoQCorAerfV1TUl1dWYgkMpM/c71/yOTVJKaWUahOfa6/l2q1b6T1jBsbX1+kxxteX3jNnEr41A59rr73kc+p0n1JKqTbz8PYm5MlleAYFUrrhd9gqKhz7jL8/gT9IIXjhwo47X4e15IQx5sfGmE+MMXnGmLeMMb7GmKeNMceNMQfsr3taqHu3MeYzY8wXxpifujJOpZRS7VP58YFvE5SnJ1A/5Vd5IKdDz+OyJGWM6Q/8CIgRkWGAJ/CAffdKERlhf73npK4n8FtgChAJJBpjIl0Vq1JKqbazVVZS8fe/A2B8fQiYdBfGxweAig8/xFZV1WHncvU1KQvgZ4yxAP5AcRvrjQK+EJHDIlID/Ddwv4tiVEop1Q5ns7KQmho8g4MZ9MYbDFi5kkFvvoFncDBSU8O5rKwOO5fLkpSIHAdWAIXACaBMRDLtuxcbY/5pjNlojLnaSfX+wJeNtovsZUoppTqZtfgEAXffzXV/+hN+N90EgN9NN3Hdn94j4O67qTne1vFI64yIdFhjTRquTz7vArOAb4B3gN8DO4DTgADPAt8RkQfPqzsDmCwiP7BvzwFGichDTs4zH5gPMHDgwJHHjh1zyedRSinlOsaYfSISc365K6f74oAjInJKRKzAFmC0iJSISJ2I2ID11E/tna8IuKbR9gBamCoUkXUiEiMiMcHBwR38EZRSSnUmVyapQuA2Y4y/McYAE4ECY8x3Gh2TAOQ5qfsRMNgYc60xxpv6Gy62uTBWpZRSbshl35MSkb8bY34P7AdqgY+BdcAGY8wI6qf7jgI/BDDGhAIbROQeEak1xiwG/kz9XYEbReQTV8WqlFLKPbnsmlRniImJkezs7M4OQymlVDt1xjUppZRS6pJoklJKKeW2NEkppZRyW5qklFJKuS1NUkoppdyWJimllFJuS5OUUkopt6VJSimllNvSJKWUUsptXTHLx1utVoqKiqjqwMW4lPvz9PSkd+/eBAUF4eGh/yZTqqu5YpJUUVERAQEBhIWFUf+8W9XdiQhWq5WSkhKKiooYOHBgZ4eklGqnK+afllVVVQQGBmqCuoIYY/D29qZ///6cO3eus8NRSl2EKyZJAZqgrlA6zadU16X/9yqllHJbmqS6ubCwMFasWNHZYSil1EXRJHURqmvrOjuETnPq1CkmT55MaGgoPj4+XHPNNSxatIiysrLODk0p1Q1pkmqno6fPcfPyHRz7+sq8EO/h4UFCQgLbt2/n4MGDpKamsmvXLubNm9fZoSmlOkldrdVlbWuSaqcVmZ9Raa1jxZ8/c/m5Nm3aRGBgINXV1U3Kk5KSmDZtGgCvvvoqEREReHt7ExERwfr16y/YZmFhIQkJCQQEBBAQEMD06dMpKioC4ODBgxhjyM3NbVJn3bp1BAUFYbVaCQwMZMGCBYwcOZJBgwYxceJEFi5cyN69ezvwkyuluoqK8jLWzp9DRblrZlM0SbXD0dPn2JFfgghk5pe4fDQ1Y8YMbDYbW7dudZSVlZWRnp5OSkoK6enpLF68mKVLl5KXl8eSJUtYuHAh27dvd9qeiBAfH09JSQm7d+9mz549FBcXEx8fj4gwZMgQYmJiSEtLa1IvLS2NWbNm4eXl1azN4uJitmzZwp133tmxH14p1SV88dHfqD53li8++tAl7WuSaocVmZ9RaxMAautsLh9N+fn5kZSUxMaNGx1lmzdvplevXtx7772sWLGCOXPmsHjxYoYMGcJDDz1EUlISv/zlL522t3PnTnJycti8eTO33HILMTExbN68mf3797Nr1y4AZs+ezVtvvYVI/ef88ssv2bt3L7Nnz27SVmJiIv7+/vTv35+AgABee+01F/WCUsqd5f1lJwCf2H92NE1SbdQwiqqzJ6m6yzSamjdvHjt27HBMyW3cuJHk5GQsFgsFBQWMGTOmyfFjx44lPz/faVsFBQWEhoYSFhbmKAsPDyc0NNRRJzExkeLiYsf03ebNmwkPDyc2NrZJWytXrmT//v1kZGRw+PBhli5d2kGfWCnVVVSdPUvJ4S8AKDn8OVXnznb4Oa6YxyJdqsajqAYNo6nV/xbtsvNGRUURHR1Namoq8fHxZGdn8+abbzr2O/uCcktfWhaRFvc1lPft25e4uDjS0tK44447SEtLIykpqdnxISEhhISEcMMNNxAYGMjtt9/OE088wTXXXHMxH1Mp1QVkrlvNwQ//CtT/LbTV1eHpacFWW4uHp4V1C7+Ph6en/WjDkNvGMmn+4ks6p0tHUsaYHxtjPjHG5Blj3jLG+BpjXjDGfGqM+acxJt0Y07uFukeNMbnGmAPGmGxXxtma80dRDS7naCo1NZUNGzYwZswYrr/+egCGDh1KVlZWk2OzsrKIjIx02k5kZCTHjx/n6NGjjrLDhw9TXFzcpM7s2bN555132LdvH7m5uc2m+s5ns9kAmt3goZTqXmLum47/VVdRW11D9blzWKuqsFbXP7TbWl2FtaqK6nPnqK2uwf+qq4i5L+GSz+myJGWM6Q/8CIgRkWGAJ/AAsAMYJiLDgYPAf16gmfEiMkJEYlwVZ1s4G0U1uBzXphITEzl58iRr164lJSXFUf7YY4/xxhtv8Nvf/pbPP/+c1atXk5aWxk9+8hOn7cTFxREVFUVSUhL79u0jOzubpKQkoqOjmTBhguO4hIQErFYrKSkpjBo1isGDBzv2/eEPf+D1118nLy+Po0eP8sc//pEFCxZw2223ERER4bpOUEp1uj6h/fn3X/2GmyZOwuLt4/QYi7c3N02cTPILv6FPaP9LPqerr0lZAD9jjAXwB4pFJFNEau37PwQGuDiGS9LSKKrB5RhNBQQEMHPmTLy9vZk5c6ajPD4+ntWrV7Ny5UoiIyNZtWoVa9asYerUqU7bMcaQkZFBcHAw48aNY/z48YSEhJCRkdFkGtDf35+EhARycnKajaJ8fX155ZVXGDt2LEOHDuXHP/4xU6dO5b333nPNh1dKuRWLlxcTH/wPRsV/Dy9f3yb7vHx9GRU/g4kPLsDT0vxu4Is6X4e04oSIHDfGrAAKgUogU0QyzzvsQeB/WmoCyDTGCPCqiKxzVawXcqFRVIPLcW3qxIkTPPDAA/To0aNJ+YIFC1iwYEGL9RpP7QEMHDiQjIyMVs+3adMmNm3a1Kw8Li6OuLi4NsWslOq+ig9+itW+Pp/x8ERsdVirqjhxsGNnllw53Xc1cD9wLRAK9DDGzG60/3GgFkhz3gJjRCQamAIsMsbc0cJ55htjso0x2adOnerQz9DaKKqBK0dTpaWlvP3222RmZrJkyZIOb18ppdrLWl3Fl5/8E6if3hs8KhaLtzcAhZ/kYK3puOvTrpzuiwOOiMgpEbECW4DRAMaYZOA+IEkavpBzHhEptv/8CkgHRrVw3DoRiRGRmODg4A79AG0ZRTVw1bWp6OhofvCDH/D8888zbNiwDm9fKaXa62jOfuqsVnr0vppZT/2CqT/+KbOe+gU9el9NndXK0Zz9HXYuV96CXgjcZozxp366byKQbYy5G/h/wJ0iUuGsojGmB+AhImfs7ycBy10YazOVNXX8KfckXp4GH0vrudxmE97LO8mvaurw8/Zs9fi2On+6TimlOlv56VMMiR3L5B/+CG8/fwBCIoYwd+WrZK57mfJTX3XYuVx5TervxpjfA/upn9b7GFgHfAL4ADvsF+s/FJEFxphQYIOI3AP0A9Lt+y3AZhF531WxOuPn7cnf/msC1VZbm+v4eHl0aIJSSil3NPKe+xl5z/3Nyn38/Zm69Kcdei6XfplXRJ4Cnjqv2Ol9yvbpvXvs7w8DUa6MrS36Bvi2fpBSSimX0cciKaWUcluapJRSSrktTVJKKaXcliYppZRSbkuTVDcXFhbGihUrOjuMFrl7fEqpzqVJ6mLUXrlP+z516hSTJ08mNDQUHx8frrnmGhYtWkRZ2YWXjn766aedfhn5o48+YuHCha4KVynVxWmSaq+vD8Evr4XSw50dSafw8PAgISGB7du3c/DgQVJTU9m1axfz5s27qPaCg4Px9/fv4CiVUt2FJqn22v0zsFbArmddfqpNmzYRGBjYbJ2mpKQkpk2bBsCrr75KREQE3t7eREREsH79+gu2WVhYSEJCAgEBAQQEBDB9+nTHqr8HDx7EGENubm6TOuvWrSMoKAir1UpgYCALFixg5MiRDBo0iIkTJ7Jw4ULHSr7OpKam8swzz/DJJ59gjMEYQ2pqKtB8us8Yw7p165gxYwY9evQgPDy8ySKPEyZMYPHipouolZeX4+/vz5YtWy742ZVSXY8mqfb4+hB89h4g9T9dPJqaMWMGNpuNrVu3OsrKyspIT08nJSWF9PR0Fi9ezNKlS8nLy2PJkiUsXLiQ7du3O21PRIiPj6ekpITdu3ezZ88eiouLiY+PR0QYMmQIMTExpKU1feZvWloas2bNwsur+aP3i4uL2bJlC3feeWeLn2PWrFk88sgjXH/99Zw4cYITJ04wa9asFo9fvnw5999/Pzk5OcyaNYsHH3yQY8eOAfULQG7evLlJ4n7rrbfo2bNni0uUKKW6MBHpNq+RI0dKS/Lz81vc12Zvf1/kmT4iT/USefrq+m0XW7RokUyePNmxvWbNGunXr59YrVYZPXq0zJ07t8nxycnJMmbMGMf2oEGD5IUXXhARkczMTPHw8JAjR4449h86dEiMMbJjxw4REXnppZdk4MCBYrPZRESksLBQjDHywQcfNDnPAw88IH5+fgLIfffdJxUVFRf8HE899ZTceOONzcobxyciAshPf/pTx7bVahU/Pz954403RESkqqpKAgMD5a233nIcM2rUKHnkkUcueP4O+f0rpVwGyBYnf9d1JNVWDaMom329Rqm7LKOpefPmsWPHDseU3MaNG0lOTsZisVBQUMCYMWOaHD927Fjy8/OdtlVQUEBoaChhYWGOsvDwcEJDQx11EhMTKS4udkzfbd68mfDwcGJjY5u0tXLlSvbv309GRgaHDx9m6dKljn09e/Z0vC601lVLhg8f7nhvsVgIDg7mq6/qH1jp4+PDnDlz2LhxIwD5+fn84x//4MEHH2z3eZRS7s+lz+7rVnb/7NsE1aDOWn9tasZrLjttVFQU0dHRpKamEh8fT3Z2dpNrNI1X1L1QGdSPmlva11Det29f4uLiSEtL44477iAtLY2kpKRmx4eEhBASEsINN9xAYGAgt99+O0888QTXXHMNBw4ccBzXq1ev9nxcgGbTisYYbLZvH/T7gx/8gOHDh1NYWMjvfvc7YmNjiYyMbPd5lFLuT0dSbXH+KKrBZRxNpaamsmHDBsaMGcP1118PwNChQ8nKympybFZWVot/sCMjIzl+/HiT5T8OHz5McXFxkzqzZ8/mnXfeYd++feTm5jZbQv58DQmk4TpRRESE49W3b18AvL29qaura98Hb8GNN97Irbfeyvr163nzzTd1FKVUN9amJGWMCXJ1IG7N2SiqQcNoyoUSExM5efIka9euJSUlxVH+2GOP8cYbb/Db3/6Wzz//nNWrV5OWlsZPfvITp+3ExcURFRVFUlIS+/btIzs7m6SkJKKjo5kwYYLjuISEBKxWKykpKYwaNYrBgwc79v3hD3/g9ddfJy8vj6NHj/LHP/6RBQsWcNtttxER4fQB90D9XXzHjh1j//79nD59utkdi+01b948fvWrX3Hu3LkL3oShlOraLpikjDFTjTGngFxjTJExZvRlist9tDSKanAZRlMBAQHMnDkTb29vZs6c6SiPj49n9erVrFy5ksjISFatWsWaNWtavMvNGENGRgbBwcGMGzeO8ePHExISQkZGRpNpQH9/fxISEsjJyWk2ivL19eWVV15h7NixDB06lB//+MdMnTqV995774Kf4bvf/S733HMPEydOJDg4mLfeeusSeqT+jsGG/ggICLiktpRS7suI89Xb63ca809gpoh8aoy5FfiViLR8r3Eni4mJkezsbKf7CgoKGDp0aPsbfWcuFGxrOUkBGE+IvN+l16amTJnCgAEDWv0e1JWiuLiYgQMH8r//+7/Nbh5x5qJ//0qpy8IYs09EYs4vb+3GiVoR+RQcK+1eWf9kbW0U1aDxaKpPeIeGUFpays6dO8nMzCQnJ6dD2+6KrFYrJ06c4PHHH+fmm29uU4JSSnVdrSWpvsaYh1vaFpFfuyYsN3Gha1Hnc9GdftHR0ZSWlvL88887ffbdleavf/0r48ePZ/Dgwbz99tudHY5SysVaS1LrgYAWtlueJ+wOaiogfyt4eoGnT+vHSx0UbK2v591xz6JrfCeegnHjxnGhKWqlVPdywSQlIs+0tM8Yc0vHh+NGvP3h4QKorWp7HYtvhyYopZS60rXry7zGmEjgASARKAOaXeTqVgL6dXYESil1RWs1SRljBlGflBKBWmAQECMiR10bmlJKqStda9+T+gB4D/ACviciI4EzbU1QxpgfG2M+McbkGWPeMsb4GmP6GGN2GGM+t/+8uoW6dxtjPjPGfGGM+Wk7P5dSSqluoLUnTpyi/kaJfkCwvaxNV62NMf2BH1E/6hoGeFI/VfhTYJeIDAZ22bfPr+sJ/BaYAkQCifapRqWUUleQCyYpEbkfuAnYDzxjjDkCXG2MGdXG9i2AnzHGAvgDxcD9wOv2/a8D8U7qjQK+EJHDIlID/Le9nlJKqStIq8/uE5EyEdkoIncBtwJPAiuNMYWt1DsOrAAKgRNAmYhkAv1E5IT9mBNAXyfV+wNfNtouspepdjp/5dvL4S9/+QvGGE6fPn1Zz6uU6n7a+oDZGGPMFuB9YD7QC3B6LalRnaupH/1cC4QCPYwxF36cdqPqTsqcTjMaY+YbY7KNMdmnTp1qY/OXxlpnvSzncVcNS8A3fr3yyiudHZZSqhtq6y3oacBjQC5ga+XYBnHAERE5BWBPcqOBEmPMd0TkhDHmO8BXTuoWAdc02h5A/VRhMyKyDlgH9c/ua2NsF620qpSp6VPZnrCdPr59XH06t7V+/Xruu+8+x/ZVV13VidEopbqrtq4ndUpEtonIERE51vBqpU4hcJsxxt/UP2J7IlAAbAOS7cckA1ud1P0IGGyMudYY4039DRfb2hirS+0u3E15TTl7Cve4/FybNm0iMDCw2bIWSUlJTJs2DYBXX32ViIgIvL29iYiIaPUBtIWFhSQkJBAQEEBAQADTp093rPp78OBBjDHk5uY2qbNu3TqCgoKwWr8dQfbu3dux8GFISAh+fn4d8ZGVUqqJtiapp4wxG4wxicaY6Q2vC1UQkb8Dv6f+potc+7nWAb8A7jLGfA7cZd/GGBNqjHnPXrcWWAz8mfrE9raIfNL+j9fxMr7IaPLTlWbMmIHNZmPr1m/zeFlZGenp6aSkpJCens7ixYtZunQpeXl5LFmyhIULF7J9+3an7YkI8fHxlJSUsHv3bvbs2UNxcTHx8fGICEOGDCEmJoa0tLQm9dLS0pg1a1aTFXOXLFlCUFAQt9xyC6+88kqTlXOVUqrDiEirL+BNIJv6u/Fes782tqXu5XyNHDlSWpKfn9/ivrb6puobuXnTzTIsdZjcvOlmKasuu+Q2W7No0SKZPHmyY3vNmjXSr18/sVqtMnr0aJk7d26T45OTk2XMmDGO7UGDBskLL7wgIiKZmZni4eEhR44ccew/dOiQGGNkx44dIiLy0ksvycCBA8Vms4mISGFhoRhj5IMPPnDUWb58uezdu1c+/vhjWbFihfj7+8uzzz7r2L9nzx4B5NSpUx3XEZeoI37/SinXAbLFyd/1tl6TihKRm1yRJN3Z0x88zY5jOxD7PRt1tjosHhasNisWDwt3vXMXnh6eABgMkwZN4qnRT3VoDPPmzSM6OpqioiIGDBjAxo0bSU5OxmKxUFBQ0Gzp9LFjx7Jtm/OZ0YKCAkJDQwkLC3OUhYeHExoaSn5+PnFxcSQmJvLoo4+yd+9e7rjjDjZv3kx4eDixsbGOOsuWLXO8HzFiBHV1dTz33HM88cQTHfrZlVKqrdN9H16JX6b9/o3fp49vH6pqqzhTc4aK2goqaysBqKytpKK2gjM1Z6iqraKPbx+Sb0xupcX2i4qKIjo6mtTUVPLy8sjOzm6SmBqvqHuhMqgfNbe0r6G8b9++xMXFOab80tLSSEpKumCMt956K+Xl5ZSUlLTpMymlVFu1NUmNBQ7YH1P0T2NMrn3V3m4t7Kow3p32Lt8d/F18PX2dHuPr6cv3hnyPLdO2EHZVmEvimDdvHqmpqWzYsIExY8Zw/fXXAzB06FCysrKaHJuVlUVkpPN/T0RGRnL8+PEmy38cPnyY4uLiJnVmz57NO++8w759+8jNzW22hPz5Dhw4gK+vL7179764D6iUUi1o63Tf3S6Nwo15e3rz+G2PE+gXyGt5r1FRW+HY52/xZ+6wuSyIWuDSGBITE3n44YdZu3Ztk+8jPfbYY8yYMYORI0cyadIk3n//fdLS0tiyZYvTduLi4oiKiiIpKYmXX34ZEeGhhx4iOjqaCRMmOI5LSEhgwYIFpKSkMGrUKAYPHuzYt337dk6ePElsbCx+fn7s2bOHJ598kvnz5+Pj04Z1t5RSqh3aNJKSRredS9tvQe9Wcr7KcSQoT1N/HaqitoKcU65f0j0gIICZM2fi7e3NzJkzHeXx8fGsXr2alStXEhkZyapVq1izZg1Tp0512o4xhoyMDIKDgxk3bhzjx48nJCSEjIyMJtOA/v7+JCQkkJOT02wU5eXlxZo1a4iNjWX48OGsWrWK5cuX8+KLL7rmwyulrmhGutEqpzExMZKdne10X0FBAUOHDr2oditrKxn71lhqbDX4ePowbsA4/lL0F6rrqvH28OaviX/F1+J8OrCjTJkyhQEDBrT6PSjl3KX8/pVSrmeM2ScizdYobNeih1eqD45/QI2thiC/IFZPWM2woGHknc7jod0PcbryNH8t/isTB050yblLS0vZuXMnmZmZ5OS4ftSmlFLuRJNUGxSfK2Zy2GSeGf0MPbx6ADAsaBjb47fz9N+epvis0yc2dYjo6GhKS0t5/vnnGTZsmMvOo5RS7kiTVBvMiZzDHOY0K+/p3ZMVd7r2CeON78RTSqkrTVtvQVdKKaUuO01SSiml3JYmKaWUUm5Lk5RSSim3pUlKKaWU29IkpZRSym1pkurmwsLCWLHCtbfJK6WUq2iSughSU9PZIXQqY0yzV+MH3yqlVEfRL/O2U21pKYfunsJ17/8JS58+nR1Op1m/fj333XefY/uqq67qxGiUUt2VjqTa6czOndjKyzmza5fLz7Vp0yYCAwOprq5uUp6UlMS0adMAePXVV4mIiMDb25uIiIhWH0BbWFhIQkICAQEBBAQEMH36dIqKigA4ePAgxhhyc3Ob1Fm3bh1BQUFYrVZHWe/evQkJCXG8/Pz8HPtSU1Pp2bMnf/rTn7jhhhvw9/dn2rRplJWV8fvf/57Bgwdz1VVXMWfOHCorKy+pj5RS3ZsmqXYq25Le5KcrzZgxA5vNxtatW789f1kZ6enppKSkkJ6ezuLFi1m6dCl5eXksWbKEhQsXsn37dqftiQjx8fGUlJSwe/du9uzZQ3FxMfHx8YgIQ4YMISYmxrEqb4O0tDRmzZqFl5eXo2zJkiUEBQVxyy238Morr2Cz2ZrUqa6u5sUXXyQtLY1du3aRnZ3N9773PV5//XXeffddMjIy+MMf/sCaNWs6sMeUUt2OiHSb18iRI6Ul+fn5Le5rq9pvvpGCYTdJ/vU3SMGwm6S2rOyS22zNokWLZPLkyY7tNWvWSL9+/cRqtcro0aNl7ty5TY5PTk6WMWPGOLYHDRokL7zwgoiIZGZmioeHhxw5csSx/9ChQ2KMkR07doiIyEsvvSQDBw4Um80mIiKFhYVijJEPPvjAUWf58uWyd+9e+fjjj2XFihXi7+8vzz77rGP/a6+9JoB8+umnjrJHHnlEPDw85NSpU01ivffeey+le9qsI37/SinXAbLFyd91vSZ1ASeWPUn5n//s2JbaWrBYwGoFLy8+HzceY/m2C3vdfTffWf5Mh8Ywb948oqOjKSoqYsCAAWzcuJHk5GQsFgsFBQU8+OCDTY4fO3Ys27Ztc9pWQUEBoaGhhIWFOcrCw8MJDQ0lPz+fuLg4EhMTefTRR9m7dy933HEHmzdvJjw8nNjYWEedZcuWOd6PGDGCuro6nnvuOZ544glHuY+Pj2OZe4B+/foREhJCUFBQk7L8/PyL7hulVPfnsuk+Y8z1xpgDjV7lxpilxpj/aVR21BhzoIX6R40xufbjnK9k6GJ9HpyLJbAPUlmJrbwcqahA7NdQpKICqaioL6+sxBIYSJ+53+/wGKKiooiOjiY1NZW8vDyys7ObJKbGK+peqAzqR80t7Wso79u3L3FxcY4pv7S0NJKSki4Y46233kp5eTklJSWOMoul6b9/jDFNpgsbys6fJlRKqcZclqRE5DMRGSEiI4CRQAWQLiKzGpW/C2y5QDPj7cc2W63xcvC59lqu3bqV3jNmYHydr7xrfH3pPXMm4Vsz8Ln2WpfEMW/ePFJTU9mwYQNjxoxxjFCGDh1KVlZWk2OzsrKIjIx02k5kZCTHjx9vsvzH4cOHKS4ublJn9uzZvPPOO+zbt4/c3NxmS8if78CBA/j6+tK7d++L+4BKKdWCyzXdNxE4JCLHGgpM/T/dZwITLlMMF8XD25uQJ5fhGRRI6YbfYauocOwz/v4E/iCF4IULXRpDYmIiDz/8MGvXrm3yfaTHHnuMGTNmMHLkSCZNmsT7779PWloaW7Y4z/txcXFERUWRlJTEyy+/jIjw0EMPER0dzYQJ3/4aEhISWLBgASkpKYwaNYrBgwc79m3fvp2TJ08SGxuLn58fe/bs4cknn2T+/Pn4+Pi4rhOUUleky3V33wPAW+eV3Q6UiMjnLdQRINMYs88YM9+l0bVB5ccHvk1Qnp5A/ZRf5QHXL+keEBDAzJkz8fb2ZubMmY7y+Ph4Vq9ezcqVK4mMjGTVqlWsWbOGqVOnOm3HGENGRgbBwcGMGzeO8ePHExISQkZGRpNpQH9/fxISEsjJyWk2ivLy8mLNmjXExsYyfPhwVq1axfLly3nxxRdd8+GVUlc0U39ThQtPYIw3UAzcKCIljcrXAl+IiNO/bsaYUBEpNsb0BXYAD4nI/zk5bj4wH2DgwIEjjx07dv4hQP1NA0OHDr2oz2CrrOTgrbchNTUYXx96jh/P2d17kOpqjLc3Q/7xdzxamA7sKFOmTGHAgAGtfg9KOXcpv3+llOsZY/Y5u7RzOUZSU4D95yUoCzAd+J+WKolIsf3nV0A6MKqF49aJSIyIxAQHB3do4A3OZmUhNTV4Bgcz6I03GLByJYPefAPP4GCkpoZz510X6kilpaW8/fbbZGZmsmTJEpedRyml3NHluCaVSPOpvjjgUxEpclbBGNMD8BCRM/b3k4Dlrg2zZdbiEwTcfTff+dnP8OzZAwC/m27iuj+9x4knllFzvNhl546Ojqa0tJTnn3+eYcOGuew8SinljlyapIwx/sBdwA/P29XsGpUxJhTYICL3AP2AdPt1EguwWUTed2WsFxKY/O+Q/O/Nyj179mTASytdeu7Gd+IppdSVxqVJSkQqgEAn5d93UlYM3GN/fxiIcmVsSiml3J8+u08ppZTb0iSllFLKbWmSUkop5bY0SSmllHJbmqSUUkq5LU1S3VxYWBgrVqzo7DCUUuqiaJK6CHW1V/byEsaYZq/GD75VSqmOokmqnSrP1PDaT7KoPFPT2aF0qvXr13PixAnHKzk5ubNDUkp1Q5qk2unwgVNUV9RyJOe0y8+1adMmAgMDqa6ublKelJTEtGnTAHj11VeJiIjA29ubiIiIVh9AW1hYSEJCAgEBAQQEBDB9+nSKiuqfTnXw4EGMMeTm5japs27dOoKCgrBarY6y3r17ExIS4nj5+fk59qWmptKzZ0927drFsGHD6NGjB+PHj+fIkSOX1B9KKffkytklTVLt9OnfTgBQ8MEJl59rxowZ2Gw2tm7d6igrKysjPT2dlJQU0tPTWbx4MUuXLiUvL48lS5awcOFCtm/f7rQ9ESE+Pp6SkhJ2797Nnj17KC4uJj4+HhFhyJAhxMTEOFblbZCWlsasWbOarKy7ZMkSgoKCuOWWW3jllVearbBbXV3Nz3/+czZu3Mjf/vY3vvnmGxYsWNCBvaOUcgcun10SkW7zGjlypLQkPz+/xX1tVXm2RtYs3C2/+eEuWbNot1Sdq7nkNluzaNEimTx5smN7zZo10q9fP7FarTJ69GiZO3duk+OTk5NlzJgxju1BgwbJCy+8ICIimZmZ4uHhIUeOHHHsP3TokBhjZMeOHSIi8tJLL8nAgQPFZrOJiEhhYaEYY+SDDz5w1Fm+fLns3btXPv74Y1mxYoX4+/vLs88+69j/2muvCSCffvqpo+zNN98ULy8vqaur64Beab+O+P0rpZrL+78i+c0Pd8kne49fUjtAtjj5u365Vubtkva8+SmH9n/l2LbVCR6exv7Tg9f/8wM8PL9dLPC66L6Mn31Dh8Ywb948oqOjKSoqYsCAAWzcuJHk5GQsFgsFBQU8+OCDTY4fO3Ys27Ztc9pWQUEBoaGhhIWFOcrCw8MJDQ0lPz+fuLg4EhMTefTRR9m7dy933HEHmzdvJjw8nNjYWEedZcuWOd6PGDGCuro6nnvuOZ544glHuY+Pj2OZe4DQ0FCsVivffPMNffr0udRuUUq5icazS5FjQzu8fZ3uu4Cb7xqIX4A31po6qitqsVbXUVtTP61VW12HtdpeXlOHX4A3N981sMNjiIqKIjo6mtTUVPLy8sjOzm6SmBqvqHuhMqgfNbe0r6G8b9++xMXFOab80tLSSEpKumCMt956K+Xl5ZSUOJYMw2Jp+u+fhvbPnxZUSnVdVeesfHXsDABfFZZTXWFtpUb76UjqAnr38+eBJ0aR9e7nfPrBCUeCaszi5cHQMaGM+V4EnhbX5Px58+bxq1/9itOnTzNmzBjHCGXo0KFkZWU1SVpZWVlERkY6bScyMpLjx49z9OhRx2jq8OHDFBcXN6kze/ZsHnroIebPn09ubi7vvvvuBeM7cOAAvr6+9O7d+9I+qFLKrXXG7JImqVZ4enlw5wPX4x/gzceZhVir6xz7LD4eRE8axC33XuvSGBITE3n44YdZu3Ztk+8jPfbYY8yYMYORI0cyadIk3n//fdLS0tiyZYvTduLi4oiKiiIpKYmXX34ZEeGhhx4iOjqaCRMmOI5LSEhgwYIFpKSkMGrUKAYPHuzYt337dk6ePElsbCx+fn7s2bOHJ598kvnz5+Pj4+O6TlBKdbqb7xpI8effUP51JbZaabKvttHfRg+LoVegX4fMLul0XxudPFzmSFDG3mu11TZOHi53+bkDAgKYOXMm3t7ezJw501EeHx/P6tWrWblyJZGRkaxatYo1a9YwdepUp+0YY8jIyCA4OJhx48Yxfvx4QkJCyMjIaDIN6O/vT0JCAjk5OcyePbtJG15eXqxZs4bY2FiGDx/OqlWrWL58OS+++KJrPrxSym00zC5Fjg3F4u08fVi8PLhxbH8eWDaK3v38L/mcpv6miu4hJiZGsrOzne4rKChg6NChF9WutaaO3z28l7paG55eHlw7PIgj/zxNndWGp8WDH/z6dizenpcSequmTJnCgAEDWv0elHLuUn7/SqnmPvrjkQ6dXTLG7BORmPPLdbqvDb78pJS6Whv+vby5Z+Fw+oX1ouRoOe+t+ScV5TUU5pcSPiLYJecuLS1l586dZGZmkpOT45JzKKVUe50/uyQ218wuaZJqgzOlVVw3si8T5tyAt299l/UL60XSM7ex+81POfN1lcvOHR0dTWlpKc8//zzDhg1z2XmUUqqtrDV1HP/sG4Bms0vHP/sXtTV1HTa7pEmqDaImXkPUxGualXv7Wbh7nmsTx9GjR13avlJKtdflnF3SJKWUUqpdLufskiYppZRS7XI5Z5dcdgu6MeZ6Y8yBRq9yY8xSY8zTxpjjjcrvaaH+3caYz4wxXxhjfuqqOJVSSrkvl42kROQzYASAMcYTOA6kA3OBlSLS4nKx9uN/C9wFFAEfGWO2iUi+q+JVSinlfi7Xl3knAodE5Fgbjx8FfCEih0WkBvhv4H6XRaeUUsotXa4k9QDwVqPtxcaYfxpjNhpjrnZyfH/gy0bbRfYypZRSVxCXJyljjDcwDXjHXrQWuI76qcATgLPn6Th7VLfTR2MYY+YbY7KNMdmnTp269IC7mbCwMFasaHFmVSml3NrlGElNAfaLSAmAiJSISJ2I2ID11E/tna8IaHzryACg2FnjIrJORGJEJCY42DVPfThfXW3HP46+KzHGNHs1fvCtUkp1lMuRpBJpNNVnjPlOo30JQJ6TOh8Bg40x19pHYg8Azlfyu8wqystYO38OFeVlnR1Kp1q/fj0nTpxwvJKTkzs7JKVUN+TSJGWM8af+Dr3Ga0f8yhiTa4z5JzAe+LH92FBjzHsAIlILLAb+DBQAb4vIJ66Mta2++OhvVJ87yxcffejyc23atInAwECqq6ublCclJTFt2jQAXn31VSIiIvD29iYiIqLVB9AWFhaSkJBAQEAAAQEBTJ8+naKiIgAOHjyIMYbc3NwmddatW0dQUBBW67cjyN69exMSEuJ4+fn5OfaVlZUxZ84c+vbti6+vL+Hh4bz00kuX0hVKqSuUS5OUiFSISKCIlDUqmyMiN4nIcBGZJiIn7OXFInJPo+PeE5EhInKdiDznyjjbI+8vOwH4xP7TlWbMmIHNZmPr1q2OsrKyMtLT00lJSSE9PZ3FixezdOlS8vLyWLJkCQsXLmT79u1O2xMR4uPjKSkpYffu3ezZs4fi4mLi4+MREYYMGUJMTIxjVd4GaWlpzJo1Cy8vL0fZkiVLCAoK4pZbbuGVV15psuLuE088QW5uLn/4wx/49NNP2bhxI/37630vSqn20ydOtEPV2bOUHP4CgJLDn1N17iy+PXq67Hx+fn4kJSWxceNGxzpSmzdvplevXtx7773ceeedzJkzh8WLFwMwZMgQ9u3bxy9/+Uuna0rt3LmTnJwcDh065FiZd/PmzURERLBr1y7i4uKYPXs2v/71r/n5z3+OMYYvv/ySvXv38otf/MLRzvLlyxk/fjw9e/Zk165dPPLII5w+fZonnngCgGPHjnHzzTczalT95caGcymlVHtpkrqAzHWrOfjhX2m4sdBWV4enpwVbbS0enhbWLfw+Hp4NT/o1DLltLJPmL+7QGObNm0d0dDRFRUUMGDCAjRs3kpycjMVioaCgoMnS8QBjx45l2zbnl+8KCgoIDQ1tkjTCw8MJDQ0lPz+fuLg4EhMTefTRR9m7dy933HEHmzdvJjw8nNjYWEedZcuWOd6PGDGCuro6nnvuOUeS+o//+A++973vsX//fu666y6mTp3KnXfe2YG9opS6UujKvBcQc990/K+6itrqGqrPncNaVYW1uv7BidbqKqxVVVSfO0dtdQ3+V11FzH0JHR5DVFQU0dHRpKamkpeXR3Z2dpPE1HhF3QuVQf10X0v7Gsr79u1LXFycY8ovLS2NpKSkC8Z46623Ul5eTklJCVC/QOOxY8d49NFHOX36NPfeey9z585t/cMqpdR5NEldQJ/Q/vz7r37DTRMnYfH2cXqMxdubmyZOJvmF39An1DXXXebNm0dqaiobNmxgzJgxXH/99QAMHTqUrKysJsdmZWURGRnptJ3IyEiOHz/eZPmPw4cPU1xc3KTO7Nmzeeedd9i3bx+5ubnNlpA/34EDB/D19aV3796OsqCgIObMmUNqaiq/+93veP3115vdAKKUUq3R6b5WWLy8mPjgf+B/VW8+2vYu1qpvH0Hv5evLLdO+S+x3E10aQ2JiIg8//DBr165t8n2kxx57jBkzZjBy5EgmTZrE+++/T1paGlu2bHHaTlxcHFFRUSQlJfHyyy8jIjz00ENER0czYcIEx3EJCQksWLCAlJQURo0axeDBgx37tm/fzsmTJ4mNjcXPz489e/bw5JNPMn/+fHx86hP5k08+SXR0NDfeeCO1tbVs2bKF8PBwx36llGorHUm1UfHBTx0JynjUX4eyVlVx4uBnLj93QEAAM2fOxNvb23EDBUB8fDyrV69m5cqVREZGsmrVKtasWeP0pgmon9LLyMggODiYcePGMX78eEJCQsjIyGgyDejv709CQgI5OTnNRlFeXl6sWbOG2NhYhg8fzqpVq1i+fDkvvvjtg0N8fHx4/PHHiYqKYsyYMZw5c6bFOw6VUupCjIjTpw11STExMZKdne10X0FBAUOHDr2odq3VVfw2JZE6qxWLtzfh0aM4vP8f1NbU4OnlxaKN/41XC9OBHWXKlCkMGDCg1e9BKecu5fevlHI9Y8w+EYk5v1yn+9rgaM5+6qxWevS+mvjHlhESMYSTXxwk44VnOffNvzias5/Bt8S23tBFKC0tZefOnWRmZpKTk+OScyillLvSJNUG5adPMSR2LJN/+CO8/fwBCIkYwtyVr5K57mXKT33lsnNHR0dTWlrK888/z7BhHbvipVJKuTtNUm0w8p77GXlP8+WsfPz9mbrUtYsGN74TTymlrjR644RSSim3pUlKKaWU29IkpZRSym1pklJKKeW2NEkppZRyW5qklFJKuS1NUt1cWFgYK1as6OwwlFLqomiSughSa2v9oG7MGNPs1fjBt0op1VH0y7ztVHe2hpMv7iPkkZF49vTu7HA6zfr167nvvvsc21dddVUnRqOU6q50JNVOlflfI5W1VOWXuvxcmzZtIjAwsNk6TElJSUybNg2AV199lYiICLy9vYmIiGj1AbSFhYUkJCQQEBBAQEAA06dPp6ioCICDBw9ijCE3N7dJnXXr1hEUFITVanWU9e7dm5CQEMfLz8/PsS8lJYUbb7yRyspKAOrq6hg7dmyTpKaUUm2hSaqdKrLrV589t++ky881Y8YMbDYbW7dudZSVlZWRnp5OSkoK6enpLF68mKVLl5KXl8eSJUtYuHBhi8tiiAjx8fGUlJSwe/du9uzZQ3FxMfHx8YgIQ4YMISYmxrEqb4O0tDRmzZqFl5eXo2zJkiUEBQVxyy238Morr2CzfTsF+vLLL2O1Wnn00UcBeO655/jiiy/YuHFjR3aPUupKICLd5jVy5EhpSX5+fov72qruXI18+V975cv/93/y5X/tlboK6yW32ZpFixbJ5MmTHdtr1qyRfv36idVqldGjR8vcuXObHJ+cnCxjxoxxbA8aNEheeOEFERHJzMwUDw8POXLkiGP/oUOHxBgjO3bsEBGRl156SQYOHCg2m01ERAoLC8UYIx988IGjzvLly2Xv3r3y8ccfy4oVK8Tf31+effbZJnF89NFH4uXlJcuWLROLxSLvvfdex3TIReqI379SynWAbHHyd91l16SMMdcD/9OoKBx4EugPTAVqgEPAXBH5xkn9o8AZoA6oFSfrjLjav7YcpCL3a2hYcssmGE+D1NX/PPHzf4CHfbFAA/43BXH19MEttncx5s2bR3R0NEVFRQwYMICNGzeSnJyMxWKhoKCABx98sMnxY8eOZdu2bU7bKigoIDQ0lLCwMEdZeHg4oaGh5OfnExcXR2JiIo8++ih79+7ljjvuYPPmzYSHhxMb++1SJMuWLXO8HzFiBHV1dTz33HM88cQTjvKYmBgef/xxnn76aRYuXMiUKVM6qEeUUlcSl033ichnIjJCREYAI4EKIB3YAQwTkeHAQeA/L9DMeHsblz1BAfS8fQCePbwQax1SVYvU1CE19dNaUmOr366qRax1ePb0ouft/Ts8hqioKKKjo0lNTSUvL4/s7OwmianxiroXKoP6UXNL+xrK+/btS1xcnGPKLy0tjaSkpAvGeOutt1JeXk5JSUmTc2VlZeHp6cmhQ4eQbrS4plLq8rlc16QmAodE5JiIZIpIrb38Q2DAZYqh3byC/em3NJoeo0IwXs67ynh50GNUCP2WROMV7O+SOObNm0dqaiobNmxgzJgxXH/99QAMHTqUrKysJsdmZWURGRnptJ3IyEiOHz/eZPmPw4cPU1xc3KTO7Nmzeeedd9i3bx+5ubnNlpA/34EDB/D19aV3796Osl//+tfs37+f//u//+PDDz9k9erV7fzUSinF5bkmBWwEFjsp3w7MbqHOEWA/sA+Y35bzuPKaVNnOY1K07K/116Psr6JlWVK289gltdsW5eXl0qNHD/H29paNGzc6ytPT08VischvfvMbOXjwoLz88stisVhk27ZtjmMaX5Oy2Wxy8803y+jRoyU7O1s++ugjue2222TkyJGOa1AiIufOnZOePXtKVFSUjBo1qkks27Ztk3Xr1klubq588cUXsn79eunVq5f86Ec/chxz4MAB8fHxkd///vciIrJ582bx9fWV3Nxcl/RPW+g1KaVcyFp1yU3QwjWpy5GgvIHTQL/zyh+nfvrPtFAv1P6zL5AD3NHCcfOBbCB74MCBLXbApf6R+up3ud8mqP/8NlGd2nh5/vDOnTtXevbsKWfPnm1SvnbtWrnuuuvEYrHIddddJ+vWrWuyv3GSEhE5duyY3H///dKzZ0/p2bOnxMfHy5dfftnsfHPmzBFAXn755Sblf/rTn2TEiBHSs2dP8ff3l2HDhslLL70kVmv9TSSVlZVy4403NruhY/bs2TJ8+HCpqrr0/5gvhiYppVzk9BciP/uOyNeHLqmZlpKUERdfKzDG3A8sEpFJjcqSgQXARBGpaEMbTwNnReSCz/eJiYmR7Oxsp/sKCgoYOnRoe0J3sNXUUbz8b1ArGC8PfG/oQ1VBaf2TJyyG/k/FYrw8L6rttpoyZQoDBgxo9XtQyrlL+f0rpS7gnbnwSTrcmAAzXrvoZowx+8TJ/QeX44kTicBbjQK5G/h/wJ0tJShjTA/AQ0TO2N9PApZfhlidqj74L6gVPAK8Cfr3SLyvCaDmyzOc3vQJtjNWqg7+C78bg1xy7tLSUnbu3ElmZiY5OTkuOYdSSl2Urw/BZ+8BUv+z9DD0Ce/QU7g0SRlj/IG7gB82Kv4N4APssN9R9qGILDDGhAIbROQeoB+Qbt9vATaLyPuujPVCar+pxm94EFd/dzAePvVd5n1NACGPxPCvLZ9T+6/qVlq4eNHR0ZSWlvL8888zbNgwl51HKaXabffPwGa/D67OCruevaTRlDMuTVL2kVLgeWURLRxbDNxjf38YiHJlbO0RMLY/9V/vasrD10Lgv7l2CqnxnXhKKeU2GkZRDUlK6lwymtLHIimllGq/xqOoBg2jqQ6kSUoppVT7nD+KatB4NNVBrqgk5eo7GZV70t+7Uh3M2SiqQQePpq6YJOXp6dlkqQl15aisrGzyBHel1CVoaRTVoINHU1dMkurduzclJSVNlpRQ3ZuIUFFRwfHjx+nbt29nh6NU93ChUVSDDhxNXTEr8wYFBVFUVMRnn33W2aGoy8jLy4t+/frRq1evzg5Fqa6vtVFUgw680++KSVIeHh4MHDiws8NQSqmuqy2jqAYd9L2pKyZJKaWUugQ1FZC/FTy9wNOn9eOlDgq21tfzvvgVIjRJKaWUap23PzxcALVVba9j8b2kBAWapJRSSrVVQL/Lfsor5u4+pZRSXY8mKaWUUm7L5etJXU7GmFPAMSe7gqhfeFG1j/bbxdO+uzjabxenO/TbIBEJPr+wWyWplhhjsp0tpqUuTPvt4mnfXRztt4vTnftNp/uUUkq5LU1SSiml3NaVkqTWdXYAXZT228XTvrs42m8Xp9v22xVxTUoppVTXdKWMpJRSSnVB3SpJGWOuMcbsMcYUGGM+McYssZf3McbsMMZ8bv95dWfH6m6MMb7GmH8YY3LsffeMvVz7rg2MMZ7GmI+NMX+wb2u/tcIYc9QYk2uMOWCMybaXab+1gTGmtzHm98aYT+1/72K7a991qyQF1AKPiMhQ4DZgkTEmEvgpsEtEBgO77NuqqWpggohEASOAu40xt6F911ZLgIJG29pvbTNeREY0un1a+61tVgHvi8gNQBT1/+11y77rVklKRE6IyH77+zPU/+L6A/cDr9sPex2I75QA3ZjUO2vf9LK/BO27VhljBgD3AhsaFWu/XRztt1YYY3oBdwC/AxCRGhH5hm7ad90qSTVmjAkDbgb+DvQTkRNQn8gAXabVCfuU1QHgK2CHiGjftc1LwE+Axss+a7+1ToBMY8w+Y8x8e5n2W+vCgVPAa/Yp5g3GmB50077rlknKGNMTeBdYKiLlnR1PVyEidSIyAhgAjDLGDOvkkNyeMeY+4CsR2dfZsXRBY0QkGphC/dT8HZ0dUBdhAaKBtSJyM3CObjK150y3S1LGGC/qE1SaiGyxF5cYY75j3/8d6kcKqgX2qYO/AHejfdeaMcA0Y8xR4L+BCcaYN9F+a5WIFNt/fgWkA6PQfmuLIqDIPtMB8Hvqk1a37LtulaSMMYb6edoCEfl1o13bgGT7+2Rg6+WOzd0ZY4KNMb3t7/2AOOBTtO8uSET+U0QGiEgY8ACwW0Rmo/12QcaYHsaYgIb3wCQgD+23VonISeBLY8z19qKJQD7dtO+61Zd5jTFjgb1ALt9eH/gv6q9LvQ0MBAqBGSJS2ilBuiljzHDqL7Z6Uv+Pl7dFZLkxJhDtuzYxxowDHhWR+7TfLswYE0796Anqp682i8hz2m9tY4wZQf2NOt7AYWAu9v9v6WZ9162SlFJKqe6lW033KaWU6l40SSmllHJbmqSUUkq5LU1SSiml3JYmKaWUUm7L0tkBKKXqGWPqqP/6RIN4IIz677scBnyB/xaRZ4wx/sB6YDhggG+Auxs9f1GpbkGTlFLuo9L+WCoH+zMo99q/e9UDOGBfDmQSUCIiN9mPux6wXuZ4lXI5TVJKdREics4Ysw+4DvgOcKzRvs86LTClXEivSSnlPvzsCwAeMMakn7/T/jSG24BPgI3A/zPG/M0Y8zNjzODLHaxSl4OOpJRyH82m++xuN8Z8TP2jvn4hIp+A49FCk6h/zuJHxphYESlwUl+pLkuTlFLub6+I3Hd+of0miS3AFmOMDbiHpqsDK9Xl6XSfUl2QMWaMMeZq+3tvIJJG16iU6i50JKVU13QdsNa+PI0H8Efq11FTqlvRp6ArpZRyWzrdp5RSym1pklJKKeW2NEkppZRyW5qklFJKuS1NUkoppdyWJimllFJuS5OUUkopt6VJSimllNv6/wcM4WGB8/8MAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(1,1, figsize=(6,4))\n",
    "plt.rcParams.update({'font.size': 14})\n",
    "\n",
    "for index, row in data.iterrows():\n",
    "    marker = '^' #'^' #'8'\n",
    "    if row['model'].startswith('yolov3'):\n",
    "        marker = '^'\n",
    "    elif row['model'].startswith('yolov5'):\n",
    "        marker= '*'\n",
    "        \n",
    "    ax.scatter(row['FPS'], row['mAP_COCO']*100, label=row['model'], marker=marker, s=150, zorder=1)\n",
    "    \n",
    "ax.set_xlabel('FPS')\n",
    "ax.set_ylabel('mAP')\n",
    "plt.legend()\n",
    "plt.tight_layout()\n",
    "\n",
    "plt.savefig('img_yolo.png', bbox_inches='tight', dpi=600)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "c5836c8b",
   "metadata": {},
   "outputs": [],
   "source": [
    "columns = [\n",
    "    'model', 'precision', 'recall', 'mAP_05', 'mAP_05_095', 'inference', 'FPS'\n",
    "]\n",
    "table = pd.DataFrame(columns = columns)\n",
    "for folder in os.listdir('yolo/yolov5/runs/train'):   \n",
    "    results = pd.read_csv('yolo/yolov5/runs/train/'+folder+'/results.csv')\n",
    "    row = [{\n",
    "        'model': folder,\n",
    "        'precision': float(results.iloc[49].values[4])*100.,\n",
    "        'recall': float(results.iloc[49].values[5])*100.,\n",
    "        'mAP_05': float(results.iloc[49].values[6])*100.,\n",
    "        'mAP_05_095': float(results.iloc[49].values[7])*100.,\n",
    "        'inference': data[data['model']==folder].values[0][3]*100,\n",
    "        'FPS': data[data['model']==folder].values[0][4]\n",
    "    }]\n",
    "    df = pd.DataFrame(row)\n",
    "    table = pd.concat([table, df])\n",
    "table.to_csv('yolo_results.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "bb5c239a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>model</th>\n",
       "      <th>precision</th>\n",
       "      <th>recall</th>\n",
       "      <th>mAP_05</th>\n",
       "      <th>mAP_05_095</th>\n",
       "      <th>inference</th>\n",
       "      <th>FPS</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>yolov3</td>\n",
       "      <td>99.351</td>\n",
       "      <td>99.555</td>\n",
       "      <td>99.489</td>\n",
       "      <td>88.625</td>\n",
       "      <td>3.12</td>\n",
       "      <td>32.0075</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>yolov3-tiny</td>\n",
       "      <td>94.863</td>\n",
       "      <td>94.971</td>\n",
       "      <td>98.633</td>\n",
       "      <td>71.984</td>\n",
       "      <td>1.56</td>\n",
       "      <td>64.0144</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>yolov5l</td>\n",
       "      <td>99.47</td>\n",
       "      <td>99.612</td>\n",
       "      <td>99.496</td>\n",
       "      <td>87.571</td>\n",
       "      <td>3.12</td>\n",
       "      <td>32.0077</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>yolov5m</td>\n",
       "      <td>99.171</td>\n",
       "      <td>99.518</td>\n",
       "      <td>99.486</td>\n",
       "      <td>85.527</td>\n",
       "      <td>1.56</td>\n",
       "      <td>64.0172</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>yolov5n</td>\n",
       "      <td>95.884</td>\n",
       "      <td>96.134</td>\n",
       "      <td>99.067</td>\n",
       "      <td>75.859</td>\n",
       "      <td>1.56</td>\n",
       "      <td>64.0166</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         model precision  recall  mAP_05 mAP_05_095 inference      FPS\n",
       "0       yolov3    99.351  99.555  99.489     88.625      3.12  32.0075\n",
       "0  yolov3-tiny    94.863  94.971  98.633     71.984      1.56  64.0144\n",
       "0      yolov5l     99.47  99.612  99.496     87.571      3.12  32.0077\n",
       "0      yolov5m    99.171  99.518  99.486     85.527      1.56  64.0172\n",
       "0      yolov5n    95.884  96.134  99.067     75.859      1.56  64.0166"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "table.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "eb1cc1e6",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\virgi\\AppData\\Local\\Temp\\ipykernel_17664\\168987356.py:2: FutureWarning: In future versions `DataFrame.to_latex` is expected to utilise the base implementation of `Styler.to_latex` for formatting and rendering. The arguments signature may therefore change. It is recommended instead to use `DataFrame.style.to_latex` which also contains additional functionality.\n",
      "  tf.write(table.to_latex())\n"
     ]
    }
   ],
   "source": [
    "with open('results_by_model.tex', 'w') as tf:\n",
    "     tf.write(table.to_latex())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "906c1a1a",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}