notebooks/08_assign_label.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x600 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import datumaro as dm\n",
    "import numpy as np\n",
    "import cv2\n",
    "\n",
    "def _make_mask(inst_id: int) -> np.ndarray:\n",
    "    mask = np.zeros([100, 100, 1], dtype=np.uint8)\n",
    "    offset = 50 * (inst_id % 2)\n",
    "    mask = cv2.putText(mask, f\"{inst_id}\", (10 + offset, 100 - 10 - offset,), cv2.FONT_HERSHEY_SIMPLEX, 1, 255)\n",
    "    return np.squeeze(mask, axis=-1).astype(bool)\n",
    "\n",
    "# Create a synthetic dataset, you may actually import the dataset as dm.Dataset.import_from().\n",
    "dataset = dm.Dataset.from_iterable(\n",
    "    [\n",
    "        dm.DatasetItem(\n",
    "            id=item_id,\n",
    "            media=dm.Image(data=np.zeros([100, 100, 3], dtype=np.uint8)),\n",
    "            annotations=[\n",
    "                dm.Mask(\n",
    "                    id=inst_id, # Instance id can exist in id field,\n",
    "                    group=inst_id, # or in group field,\n",
    "                    image=_make_mask(inst_id),\n",
    "                    attributes={\"inst_id\": inst_id}, # or in attributes field.\n",
    "                    label=0, # We assume that the class label as 0 for all masks in this example.\n",
    "                )\n",
    "                for inst_id in range(item_id * 2, item_id * 2 + 2)\n",
    "            ],\n",
    "        )\n",
    "        for item_id in range(4)\n",
    "    ],\n",
    "    categories=[\"label_0\", \"label_1\"],\n",
    ")\n",
    "\n",
    "# Visualize it. Because the class label is 0 for all masks, colors of all masks are same (blue).\n",
    "viz = dm.Visualizer(dataset, ignored_types=[dm.AnnotationType.bbox])\n",
    "fig = viz.vis_gallery([item_id for item_id in range(4)], dm.DEFAULT_SUBSET_NAME)\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# You may have a map, {instance_id: class_id}\n",
    "map_inst_id_to_label_id = {\n",
    "    inst_id: inst_id % 2\n",
    "    for inst_id in range(8)\n",
    "}\n",
    "\n",
    "# Convert label_id mannually and export it to coco_instances\n",
    "for item in dataset:\n",
    "    for ann in item.annotations:\n",
    "        if ann.type == dm.AnnotationType.mask:\n",
    "            inst_id = ann.id\n",
    "            ann.label = map_inst_id_to_label_id[inst_id]\n",
    "\n",
    "dataset.export(\"ws_test/inst-mask\", \"coco_instances\", save_media=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/vinnamki/datumaro/datumaro/components/visualizer.py:310: UserWarning: Mask should has dtype == bool, but its dtype == uint8. Try to change it to bool dtype.\n",
      "  warnings.warn(\n",
      "/home/vinnamki/datumaro/datumaro/components/visualizer.py:267: UserWarning: bbox in self.ignored_types. Skip it.\n",
      "  warnings.warn(msg)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjgAAAH4CAYAAABHdZffAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjYuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8o6BhiAAAACXBIWXMAAA9hAAAPYQGoP6dpAABHwElEQVR4nO3deXgUVb7/8U9VJ53ubCRAAMNOkB3l/iKOCgiKgF4UQZnIOKOgZmAGB713FrzqKILjNiri7rihos5FGFxGWVWURXC5KiKKIJuIIARIyN7p7vP7I6alSQJZOlvl/XqePNCV6lOnO+lvPn3qnGrLGGMEAADgIHZDdwAAACDSCDgAAMBxCDgAAMBxCDgAAMBxCDgAAMBxCDgAAMBxCDgAAMBxCDgAAMBxCDgAAMBxCDiNyKRJkxQfH9/Q3YioSZMmqUuXLjW+/7333qtu3brJ5XJpwIABEevX0d577z1ZlqX33nuvTtoHUH9uu+02WZalrKyshu5KxJQ9ppqaN2+eevXqpejoaCUlJUWuY0fZuXOnLMvSc889Vyft10S9B5znnntOlmXpk08+CW0r++GVfcXGxqpTp0666KKLNHfuXBUXF9f6uHv27FFGRoaSkpKUmJioiy++WNu3b69Vm2vWrNEFF1yg9u3by+PxhPr88ssv17q/9aWgoEC33XZbo/zjvnz5ck2fPl2DBg3S3Llzdeedd9bbsV9++WXNmTOn3o4HVEdD1NFvvvlG//3f/62zzjpLHo9HlmVp586dtXwk0saNGzV+/Hh17txZHo9H7du314gRI/Twww/Xuu36dOedd+q1115r6G6Us3nzZk2aNElpaWl66qmn9OSTT9bbsRcvXqzbbrut3o53rKgGO3IFHn/8ccXHx6u4uFh79uzRsmXLdPXVV2vOnDl688031bFjxxq1m5eXp3POOUc5OTm66aabFB0drQceeEBDhw7V559/rlatWlW7zQULFuiyyy7TgAEDdP311ys5OVk7duzQqlWr9NRTT+nyyy+vUV/rW0FBgWbOnClJGjZsWMN25hjvvvuubNvWM888I7fbXa/Hfvnll/Xll1/qv/7rv+r1uEBt1VUdXbdunR566CH16dNHvXv31ueff17rvn7wwQc655xz1KlTJ/32t79Vu3bttHv3bq1fv14PPvigpk2bVutj1Jc777xT48eP19ixYxu6K2Hee+89BYNBPfjgg+revXu9Hnvx4sV69NFHGyzkNKqAM378eLVu3Tp0+9Zbb9VLL72kK6+8Ur/85S+1fv36GrX72GOPaevWrfroo480cOBASdIFF1ygfv366f7776/RyMBtt92mPn36aP369eX++O7fv79G/US4/fv3y+v11nu4AZqyuqqjY8aMUXZ2thISEnTfffdFJODccccdatGihT7++ONyp06oo5FR9jzW1ampxqzRz8H59a9/rczMTH344YdasWJFaHtBQYE2b95cpfOsCxcu1MCBA0PhRpJ69eql4cOH65VXXqlRv7Zt26aBAwdW+Me3TZs2of9XNr/jeOcrt2/frlGjRikuLk6pqamaNWuWjv3Q9//93/9Venq6EhISlJiYqP79++vBBx8M2yc7O1v/9V//pY4dOyomJkbdu3fXPffco2AwGOpDSkqKJGnmzJmhoe2ytF1SUqLNmzdr7969VXpOXnvtNfXr108ej0f9+vXTq6++WuF+wWBQc+bMUd++feXxeNS2bVtNmTJFhw8fDu1jWZbmzp2r/Pz8UL/Knqu5c+fq3HPPVZs2bRQTE6M+ffro8ccfL3ecox/L0bp06aJJkyZV+jiGDRumt956S7t27QoduzbziICGFok62rJlSyUkJES0X9u2bVPfvn0r/ON7dB09Xr2s7HWelZWljIwMJSYmqlWrVrr++utVVFQUts+KFSs0ePBgJSUlKT4+Xj179tRNN90Utk9xcbFmzJih7t27KyYmRh07dtT06dPDTvlZlqX8/Hw9//zzoZpxdI3ZvHmzvvvuuyo9J2vWrNHAgQPl8XiUlpamf/zjH5Xu++KLLyo9PV1er1ctW7bUhAkTtHv37tD3u3TpohkzZkiSUlJSwp6r119/XaNHj1ZqaqpiYmKUlpam22+/XYFAIOwYldXLYcOGHXfUf9KkSXr00UclKezUaX1qVCM4lbniiiv05JNPavny5RoxYoQk6aOPPtI555yjGTNmHHf4KxgM6osvvtDVV19d7nunn366li9frtzc3Gq/cDt37qx33nlH33//vTp06FCt+x5PIBDQ+eefrzPOOEN///vftXTpUs2YMUN+v1+zZs2SVPqi/NWvfqXhw4frnnvukSR9/fXXWrt2ra6//npJpYVr6NCh2rNnj6ZMmaJOnTrpgw8+0I033qi9e/dqzpw5SklJ0eOPP67f//73GjdunC655BJJ0imnnCKpdN5S7969NXHixBNOHFu+fLkuvfRS9enTR3fddZcOHjyoq666qsLnZsqUKXruued01VVX6brrrtOOHTv0yCOP6LPPPtPatWsVHR2tefPm6cknn9RHH32kp59+WpJ01llnSSodgu/bt6/GjBmjqKgo/fvf/9bUqVMVDAZ17bXX1vpncPPNNysnJ0fff/+9HnjgAUly3ORvND+1qaN1pXPnzlq3bp2+/PJL9evXL6JtZ2RkqEuXLrrrrru0fv16PfTQQzp8+LBeeOEFSdKmTZt04YUX6pRTTtGsWbMUExOjb7/9VmvXrg21EQwGNWbMGK1Zs0aTJ09W7969tXHjRj3wwAPasmVLaM7NvHnzlJmZqdNPP12TJ0+WJKWlpYXa6d27t4YOHXrCuY4bN27UyJEjlZKSottuu01+v18zZsxQ27Zty+17xx136JZbblFGRoYyMzN14MABPfzwwzr77LP12WefKSkpSXPmzNELL7ygV199NXTqsqy+P/fcc4qPj9cf//hHxcfH691339Wtt96qI0eO6N57763NUy+ptM7/8MMPWrFihebNm1fr9mrE1LO5c+caSebjjz8ObZsxY4aRZA4cOFDhfQ4fPmwkmXHjxoW2rVy50kgyM2bMOO7xDhw4YCSZWbNmlfveo48+aiSZzZs3V/txPPPMM0aScbvd5pxzzjG33HKLWb16tQkEAmH7lfVz5cqVYdt37NhhJJm5c+eGtk2cONFIMtOmTQttCwaDZvTo0cbtdoeen+uvv94kJiYav99faf9uv/12ExcXZ7Zs2RK2/X/+53+My+Uy3333nTHm5+enouexrI8TJ0484fMxYMAAc9JJJ5ns7OzQtuXLlxtJpnPnzqFtq1evNpLMSy+9FHb/pUuXlts+ceJEExcXV+5YBQUF5baNGjXKdOvWLWxbZY+rc+fOYY+pop/R6NGjw/oNNCb1XUePde+99xpJZseOHTXpfsjy5cuNy+UyLpfLnHnmmWb69Olm2bJlxufzhe1XUb0sc2z/y56HMWPGhO03depUI8ls2LDBGGPMAw88cNznyxhj5s2bZ2zbNqtXrw7b/sQTTxhJZu3ataFtcXFxldZKSWbo0KGVHqfM2LFjjcfjMbt27Qpt++qrr4zL5TJH/7neuXOncblc5o477gi7/8aNG01UVFTY9sp+Lyqqo1OmTDGxsbGmqKgotO3Yellm6NChYY+pop/RtddeG9bv+tboT1FJP797zs3NDW0bNmyYjDEnfNdRWFgoSYqJiSn3PY/HE7ZPdVx99dVaunSphg0bpjVr1uj222/XkCFDdPLJJ+uDDz6odntH+8Mf/hD6v2VZ+sMf/iCfz6e3335bUum51Pz8/LCh5mMtWLBAQ4YMUXJysrKyskJf5513ngKBgFatWnXCfnTp0kXGmBOO3uzdu1eff/65Jk6cqBYtWoS2jxgxQn369CnXrxYtWmjEiBFh/UpPT1d8fLxWrlx5wn55vd7Q/3NycpSVlaWhQ4dq+/btysnJOeH9geaoNnW0rowYMULr1q3TmDFjtGHDBv3973/XqFGj1L59e73xxhu1avvY0dyyCcuLFy+W9POclNdffz102v5YCxYsUO/evdWrV6+wenXuuedKUpXqlSQZY044ehMIBLRs2TKNHTtWnTp1Cm3v3bu3Ro0aFbbvokWLFAwGlZGREdavdu3a6eSTT652Hc3NzVVWVpaGDBkSOm3pBE0i4OTl5UlSjc7/lv0QK1oiWXY+9ugfdHWMGjVKy5YtU3Z2tlatWqVrr71Wu3bt0oUXXljjCXK2batbt25h23r06CFJoSWZU6dOVY8ePXTBBReoQ4cOobB1tK1bt2rp0qVKSUkJ+zrvvPMkRXYC365duyRJJ598crnv9ezZs1y/cnJy1KZNm3J9y8vLq1K/1q5dq/POO09xcXFKSkpSSkpK6Lw5AQeoWG3qaF0aOHCgFi1apMOHD+ujjz7SjTfeqNzcXI0fP15fffVVjds9th6lpaXJtu1QHb3ssss0aNAgZWZmqm3btpowYYJeeeWVsLCzdetWbdq0qVytKqvJkayjBw4cUGFhYZXrqDFGJ598crm+ff3111Xq16ZNmzRu3Di1aNFCiYmJSklJ0W9+8xtJzqmjTWIOzpdffilJNVri1rJlS8XExFQ4UbZsW2pqaq36FxsbqyFDhmjIkCFq3bq1Zs6cqSVLlmjixImVTqo6diJXdbRp00aff/65li1bpiVLlmjJkiWaO3eurrzySj3//POSSs8djxgxQtOnT6+wjbIXaH0LBoNq06aNXnrppQq/XzbpuTLbtm3T8OHD1atXL82ePVsdO3aU2+3W4sWL9cADD1T6TuxotXnugaaqNnW0Prjd7tBikB49euiqq67SggULNGPGjIjU0WPb8Hq9WrVqlVauXKm33npLS5cu1fz583Xuuedq+fLlcrlcCgaD6t+/v2bPnl1hmzVdcl9bwWBQlmVpyZIlcrlc5b5/ojmD2dnZGjp0qBITEzVr1iylpaXJ4/Ho008/1Q033BBWR4/33Fd07MakSQScsglKxw7TVYVt2+rfv3/YBbHKfPjhh+rWrVtE39Gcdtppkn4OT8nJyZJKf6GOVjbqcaxgMKjt27eHBZAtW7ZIUthKHrfbrYsuukgXXXSRgsGgpk6dqn/84x+65ZZb1L17d6WlpSkvLy80YlOZSMxq79y5s6TSdxXH+uabb8Jup6Wl6e2339agQYNqNHL273//W8XFxXrjjTfChnErGpJNTk4u97z7fL4qrQqr79n+QF2rTR2tb7Wto1JpPeratWvo9rfffqtgMBhWR23b1vDhwzV8+HDNnj1bd955p26++WatXLlS5513ntLS0rRhwwYNHz78hDWhtjUjJSVFXq+3ynXUGKOuXbvW6M3qe++9p4MHD2rRokU6++yzQ9t37NhRbt+K6qhU+twfe7bhWA1dRxv9KaqXX35ZTz/9tM4880wNHz48tL06yxvHjx+vjz/+OCzkfPPNN3r33Xf1y1/+skb9eueddyrcXnZ+t2xIsXPnznK5XOXmvDz22GOVtv3II4+E/m+M0SOPPKLo6OjQ4z948GDY/rZth2bGl52Ky8jI0Lp167Rs2bJy7WdnZ8vv90sqHX0q23asqi4TP+mkkzRgwAA9//zzYUObK1asKDfEnJGRoUAgoNtvv71cO36/v8J+HK3sHYM5atl8Tk6O5s6dW27ftLS0cs/7k08+WaV3fXFxcY4ZpgUiUUfrwsqVK8tdAkMqX0cTExPVunXratXRsiXKZcqujHzBBRdIkg4dOlTuPmUfB3N0Hd2zZ4+eeuqpcvsWFhYqPz8/dDsuLq7S+lWVZeIul0ujRo3Sa6+9Frbv119/Xa6OX3LJJXK5XJo5c2a5588YU+5vREXHKtu3jM/nq/D5TEtL0/r16+Xz+ULb3nzzzbDl6JWJi4uTVPHfl/rQqEZwFi5cqPj4ePl8vtAVONeuXatTTz1VCxYsCNu3Ossbp06dqqeeekqjR4/Wn//8Z0VHR2v27Nlq27at/vSnP4XtO2zYML3//vsVvuiOdvHFF6tr16666KKLlJaWpvz8fL399tv697//rYEDB+qiiy6SJLVo0UK//OUv9fDDD8uyLKWlpenNN9+s9Bypx+PR0qVLNXHiRP3iF7/QkiVL9NZbb+mmm24Knb7JzMzUoUOHdO6556pDhw7atWuXHn74YQ0YMEC9e/eWJP3lL3/RG2+8oQsvvFCTJk1Senq68vPztXHjRi1cuFA7d+5U69at5fV61adPH82fP189evRQy5Yt1a9fP/Xr169ay8TvuusujR49WoMHD9bVV1+tQ4cO6eGHH1bfvn1D5/4laejQoZoyZYruuusuff755xo5cqSio6O1detWLViwQA8++KDGjx9f6XFGjhwZGr2aMmWK8vLy9NRTT6lNmzblglhmZqZ+97vf6dJLL9WIESO0YcMGLVu2LOwiaJVJT0/X/Pnz9cc//lEDBw5UfHx86GcKNGZ1VUdzcnJCIaFsKfUjjzyipKQkJSUlhS2OmDRpkp5//nnt2LHjuNeQmjZtmgoKCjRu3Dj16tVLPp9PH3zwgebPn68uXbroqquuCu2bmZmpu+++W5mZmTrttNO0atWq0Oh2RXbs2KExY8bo/PPP17p16/Tiiy/q8ssv16mnnipJmjVrllatWqXRo0erc+fO2r9/vx577DF16NBBgwcPllS6tP6VV17R7373O61cuVKDBg1SIBDQ5s2b9corr2jZsmWh0ab09HS9/fbbmj17tlJTU9W1a1f94he/kFT1ZeIzZ87U0qVLNWTIEE2dOlV+vz9UR7/44ovQfmlpafrb3/6mG2+8UTt37tTYsWOVkJCgHTt26NVXX9XkyZP15z//udLjnHXWWUpOTtbEiRN13XXXybIszZs3r8K/e5mZmVq4cKHOP/98ZWRkaNu2bXrxxRfDlsFXJj09XZJ03XXXadSoUXK5XJowYcIJ7xcx9b1s63jLG8u+PB6P6dChg7nwwgvNs88+G7ZkrUx1lzfu3r3bjB8/3iQmJpr4+Hhz4YUXmq1bt5bbLz093bRr1+6E7f3zn/80EyZMMGlpacbr9RqPx2P69Oljbr75ZnPkyJGwfQ8cOGAuvfRSExsba5KTk82UKVPMl19+WeEy8bi4OLNt2zYzcuRIExsba9q2bWtmzJgRtvx84cKFZuTIkaZNmzbG7XabTp06mSlTppi9e/eGHTc3N9fceOONpnv37sbtdpvWrVubs846y9x3331hyzA/+OADk56ebtxud9hzWp1l4sYY869//cv07t3bxMTEmD59+phFixaZiRMnVrjc+sknnzTp6enG6/WahIQE079/fzN9+nTzww8/lHs+jvXGG2+YU045xXg8HtOlSxdzzz33mGeffbbcstVAIGBuuOEG07p1axMbG2tGjRplvv322yotE8/LyzOXX365SUpKKrfUHWhoDVFHy+pBRV/Hvj4uvfRS4/V6zeHDh4/b5pIlS8zVV19tevXqZeLj443b7Tbdu3c306ZNMz/++GPYvgUFBeaaa64xLVq0MAkJCSYjI8Ps37+/0mXiX331lRk/frxJSEgwycnJ5g9/+IMpLCwM7ffOO++Yiy++2KSmphq3221SU1PNr371q3KX1vD5fOaee+4xffv2NTExMSY5Odmkp6ebmTNnmpycnNB+mzdvNmeffbbxer3l6qaquEzcGGPef//9UD3u1q2beeKJJ0KP6Vj/+te/zODBg01cXJyJi4szvXr1Mtdee6355ptvyj0fxy4TX7t2rTnjjDOM1+s1qampoSX6x9ZCY4y5//77Tfv27U1MTIwZNGiQ+eSTT6q0TNzv95tp06aZlJQUY1lWvS8Zt4w5wVBFM5Kbm6uWLVtqzpw5EblgHAA0R23bttWVV14ZkQvGATXV6Ofg1KdVq1apffv2+u1vf9vQXQGAJmnTpk0qLCzUDTfc0NBdQTPHCA4AAHAcRnAAAIDjEHAAAIDjEHAAAIDjEHAAAIDjEHAAAIDjVPlKxg39mRKAE7BosXmjjgK1V9U6yggOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwHAIOAABwnKiG7gCarxiX5LKlgpKG7gkANDyXJXmj6/YYxX6pJFi3x2gsCDhoELYlTR/k1n+0c+mq1wuVU9zQPQKAhjWgna3HRnvq9NTKXWt8WrTZX4dHaDwIOKh3UbaU4Jb6t7HVN8VWFCdKAUDFAemHXCPbilyb3Vta6tnK1rZDRpsPBlVQYiLXeCNnGWOq9GgtK4LPOJqtKFv66xC3BnWK0n+0s3WwwOisZ/N1sLChe1Y/qvhyg0NRR3E8lhTRN3yWpIt7Remfl3r0+7eK9fyGEvmDUrCJl6Gq1lFGcFBvbEuKjZZax1pK9kb2hQwATZ1R5OfHZBWYn/4NyheIbNuNHQEH9SLGJd0wyK2SoDRrlU+S9I8LPerTmpQDAIg8Ag7qnKXSkZuBqS59nxvU4UKjoJEOFjbxcVIAaMQslY4KbTtklO9r6N7UPwIO6lRctDR1oFtZBUFNfrNIeT6jkmDpckgAQN3o38bWZX2jtXx7ic56Nl9HmuFKVc4PoE7Fuy39un+U+qS49GO+UW4zfBcBAPWtZytbk9OjVRKQDhY2n2vfHI0RHERUK6+U0Sda9k/Rec+RoMbNL1ROsWnyM/cBoLE7pa2tQR1d+mxfQKc9la8f85pv4SXgIKJSE2zddV6M3K7S23es9um1b5rZ1H0AaCBDOrl0+zkxGvxsgb7Lab7hRiLgIMK2Hgpq0LMFKrvcx/5m/O4BAOpLvza2eray9enegIbMLdC2w83wnNQxCDiIqCK/tOkALywAqC+WpAn9ojX5/0Ur/cl87T7CG0uJScYAAMCBCDgAAMBxOEUFAEAT1SHRUkqspdW7/Fq32x/6aAYQcAAAaJJclnTLELcGd3bp7LkFzeZDi6uKgAMAQBPTNs5Su3hL3Vu6lOyxNbC9S5uzyi/w2JtrVNxMr9RBwAEAoAlxu6SHzo/RiLQoxbsl25JeGe+V/5h8EzTSRf8s0Lrvm+fKVgIOAABNSCAovbHFr80Hjx9cjJGiXVLrWEtHio18zWwkh4ADAEATEjDSSxv9qspnFg/t4tKH18Rq6uJCLdvWvBIOy8QBAGiCTBW+9uYG9fb2EtmWlOCWopvRX/1m9FABAGhevjlo9PvFxdqXZ7T0N7G6qEfzOXFDwAEAwMGCpvQr/SRbKXFVObHlDAQcAADgOAQcAAAcLr/E6KM9QeWXGEXZqtIE5aaOgAMAgMNtP2R0wUsF+nxfUA+dH6OLergaukt1joADAIDDBSXll0h5PqMJ/aI1oB0BBwAAoMkh4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMdpPh8rCgBAI9I6Vjop3taWg0EVB+r2WJ4oqUcrW/Fu6erXC7X5YLBuD9gIMIIDAEADyOgTrdVXxSqtZd3/Ke7Rytaaq2LVL8Wl174JaHOWqfNjNjQCDgAADcBlS25X/XzwpaWfjtUcPmXzJwQcAADgOAQcAAAawEd7Arp7jU9x0aXzcaLr4C9ytC21jrXkjZLuXOPT//1Qx5N9GhECDgAADeDDPUHdvsqnji1srbsmXgPaRf5PcnqqrQ+viVObOEu3v+/TJ3udP7m4DKuoAABoIEbSruyg3t1eIm+0FO/++XvFfqmkmnkk2pZijvrL7nFZent7iXYfMXL+tOJwBBwAABrQJ3uD+nRxsc7u7NI7V8aGTq3ctcanRZv91WprXO8o/eXM0pQUlPSX5UX6/eJiBZtbuhEBBwCABhc0UlaB0Q+5RvZPK51iokqvX1PVcOKyJLct/ZBnQm0eLKz6/Z2GgAMAQCOwaX9QGQsKQ7e7t7T1wliPYqOrtrY732d0y8pizd/086iPv/lMuSmHgIMGYSTtOBxUort5vwABoIxR+Jyb7CKjFh5bie5K7xImypZyiqs/b8epLGNMlQavrOZ0dSDUC7erdEi1sHqnmJu0Kr7c4FDUUVSHJckbXfULARpJhSVy/GTiqtZRAg5Qjwg4zRt1FKi9qtZRroMDAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAch4ADAAAcJ6qhO1BbVlSMZFkn3M+UFNVDbwCg6alqHS1jSoolmbrrEBABTTrg2J44tfrP/5YrvuVx9wsW5yvr3/crWJBdPx0DgCaiqnW0jAn4dfCt2fJn76vjngG10/QCjuvnLlvuOAWK8k78PqKkSLYnTsHivJ+3BQLiHQiAZqkmdbRMMCDLHVvaBnUUjZhljKnSb6dVjeHLuhLVoq2Shv9Wlu2SJBV++5Hyv3xHJhg8/h0tS7G9hiiu95DS28bo8PvPy5+1q457DISr4ssNDtWk62gZS/J2TVf8qaOoo2gQVa2jTWcEx7JleeLl8iRIP70wTTAg4/dV7f5+n2xPwk83jFyeePktWzJVfFEDQFNX2zr6E1NSLNubQB1Fo9YkRnCiWnZQ4umXqGDLWhXv3hTabgJ+KeivWiO2S5YrOnTT3e5kxfU9R7mfvaWSH7dFustAhRjBad6afB0tY9myotzUUTSIqtbRJrFM3BWfrLj+w2W53DIlRaGvar0og4Hw+1pSXP/hikpsU3cdB4BGIiJ1tIwJUkfR6DWJgAMAAFAdTSbgBIvyZGryTqPSBgMKFuVJnDIA0ExQR9GcNOo5OFHJJ8nT5f+pJGuX/Dk/KliQU+3JcJVyRcsVlyRXQmu523RV0e4v5c/6LjJtA5VgDk7zRh0Fas8Rc3Dcbbsr+bzJkjEKHDkQuRelJAVKStv0FSn53N/K06Fv5NoGgEaCOormqlEHHAAAgJog4ACA03FqFM0QAQcAHKxo1wbt/9+b5Tuwo6G7AtSrpnMlYwBAtQULj6j4+00n3hFwmEYZcGxvolzxrRQoyFHWa3ep5GDdzcr3H/lRWa/fLVNSrOg2XRXIO8ynjgMA0MQ1ylNUsb0Gq93ldytQkK3CresVLMyts2OZ4gIVfvuhSrL3qk3GLMWfOqrOjgUAAOpHoxzBsSyX5Krnrhkjy46S7EaZ+QCg8XFFyxWXLMvt0eG3/6EiToWhEeGvOQCgRmJO6qGTJj0oV2wL5W1YKv/B3Q3dJSCkUY7gAAAaMVeUbHesZNnK/b9/q+TQnobuEVAOIzgAgGrxdPkPtb3iPskElbP2ZUZu0CgRcAAA1WJHexTVom1DdwM4LgIOAABwHObgAACqxrIl2yXj96lw28cKFNXdJTyA2iLgAACqxJt2muJOGamcNS8r67W7paC/obsEVIqAAwA4AUuy7dKrzMfEyxQXEG7Q6BFwAADH5e1xhjydT1Xup2+p4OvVMn5fQ3cJOCECDgCgcpat6JYd5D7pZAWL8mX8xQ3dI6BKCDgAgApY8vY4Q9EtO6hgyzrlfb5EweL8hu4UUGUsEwcAlGfb8nQeoNgeZylYkK1gUZ5kTEP3CqgyRnAAAD+zbHm7pcv2tlDuZ2/pyLr5jNygSWqUAceYwE+T2Kz6O6hlyQT8UjBYf8cEgMbGdinu1FFyeeJVsHk1c27QZFnGVG3M0bLqL2zYngS54lvKjmsh25Og4u82Klh4pE6OZcXEydP5VBlfgQL52QrkH1awIKdOjgVU8eUGh6rPOlptli1Pp1NkRbnlzzsoU5wvf/aPkvidReNS1TraKOfgBItyVZK1Sy5vC7UeM13RrTrW2bGiEtuo9UV/UVTSSSo5sJNwA6BZsqJjlDR0ouJOHamSA7vkz94nwg2askZ5igoAUE9sl9ztTpblitbhd58u/fgFLuIHB2iUIzgAgPphx8Sr9UV/VmyvwSre85X8Wd81dJeAiGAEBwCaIztKUS3by3Z7dPidp+Q/sp9l4HAURnAAoBlyxbdU2wl/k6fLf6jw2w9Vsn9HQ3cJiKhGuYqqTFRSO3k6D1DJoe/lP3JAwfzsyC1ZdEXLFZcsV0IruVt3VtH3m+Q/uDsybQOVYBVV89YoVlHZUXLFt5QrNknRbbqo5Mft8v34bUP3CqiyJr2Kqow/e5/yNiyVLEsnTZwjT+dTItZ2zEk9dNKkB+WKbaG8DUsJNwCahaiW7dXuyvsV3aaL8r9YTriBYzWZOThWTKxkR7C7tl3aZn1eTBAAGortkh0TJ9vtUe5nizklBcdr1CM4AIDIcLfrrnZXzpYrNklH1v5Tvn1bG7pLQJ1qEiM4gbxDyv9iuYzfJ8vtDc30L/1ohSper8F2yXJF/3zbSPlfLJc/58c66DEANBKWLSvKLcvlVtHOz+Q/cqChewTUi0Y9yfiYDii6TTe1HPE7WbZLkpS7YZnyNyyr0t1jew1R4unjJElGRtnvPqviPV9Lhs+eQv1hknHz1hB11NNlgFqcfQU1D45R1TraJEZwJEnGKFiUp0DeIemngGNZtqyoGBkTOP59LVtyRcmfdyi0KVCUywsdgHNZtmS7FJXcXu623WXHJUm2rXqZmRAMUl/R4JrOCE6ZoyYau+JbKmnIb2R7E457F+MrUvbqF8NPR3EpcjQARnCat/qso97upyt+wH8qulUHuRJT5D+0p95Oyed/9Z4Kvnq/Xo6F5sd5Izhljgompjhfltsr23OCgGPZChbnE2oANBuWyy3bE69AfrYC+dmSdMJaGbFjR8XUy3GA42l6IzjHsKLcpUOxx2NM5C4QCNQCIzjNW73W0WMXVtSjai0AAaqpqnW0yQccoCkh4DRv1FGg9hxxJWMAAICaIOAAAADHIeAAAADHIeAAAADHIeAAAADHIeAAAADHIeAAAADHIeAAAADHIeAAAADHIeAAAADHIeAAAADHIeAAAADHIeAAAADHIeAAAADHIeAAAADHIeAAAADHIeAAAADHsYwxpqE7AQAAEEmM4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4DQit912myzLUlZWVkN3JWLKHlNNzZs3T7169VJ0dLSSkpIi17Gj7Ny5U5Zl6bnnnquT9gHUH+poec21jtZ7wHnuuedkWZY++eST0LayH17ZV2xsrDp16qSLLrpIc+fOVXFxca2OuWjRIl122WXq1q2bYmNj1bNnT/3pT39SdnZ2rdrduHGjxo8fr86dO8vj8ah9+/YaMWKEHn744Vq1W9/uvPNOvfbaaw3djXI2b96sSZMmKS0tTU899ZSefPLJejv24sWLddttt9Xb8YDqaIg6+uqrr2rUqFFKTU1VTEyMOnTooPHjx+vLL7+sVbvU0brVnOtoVIMduQKPP/644uPjVVxcrD179mjZsmW6+uqrNWfOHL355pvq2LFjjdqdPHmyUlNT9Zvf/EadOnXSxo0b9cgjj2jx4sX69NNP5fV6q93mBx98oHPOOUedOnXSb3/7W7Vr1067d+/W+vXr9eCDD2ratGk16mtDuPPOOzV+/HiNHTu2obsS5r333lMwGNSDDz6o7t271+uxFy9erEcffZSQgyanruroxo0blZycrOuvv16tW7fWvn379Oyzz+r000/XunXrdOqpp1a7Tepo3WvOdbRRBZzx48erdevWodu33nqrXnrpJV155ZX65S9/qfXr19eo3YULF2rYsGFh29LT0zVx4kS99NJLyszMrHabd9xxh1q0aKGPP/643JDf/v37a9RPhCt7HutqSBVworqqo7feemu5bZmZmerQoYMef/xxPfHEE9Vukzpa95pzHW30c3B+/etfKzMzUx9++KFWrFgR2l5QUKDNmzdX6TzrseFGksaNGydJ+vrrr2vUr23btqlv374V/tK0adMm9P/jnZe0LKvCZJuVlaWMjAwlJiaqVatWuv7661VUVBS2z4oVKzR48GAlJSUpPj5ePXv21E033RS2T3FxsWbMmKHu3bsrJiZGHTt21PTp08OGqi3LUn5+vp5//vnQ0PakSZNC39+8ebO+++67Kj0na9as0cCBA+XxeJSWlqZ//OMfle774osvKj09XV6vVy1bttSECRO0e/fu0Pe7dOmiGTNmSJJSUlLCnqvXX39do0ePDg2Vp6Wl6fbbb1cgEAg7RpcuXcIeS5lhw4ZV+DtRZtKkSXr00UclKWzIH2iqIlFHK9KmTRvFxsbW+HQ/dbQ86mjkNKoRnMpcccUVevLJJ7V8+XKNGDFCkvTRRx/pnHPO0YwZM2o0/LVv3z5JCnunUx2dO3fWunXr9OWXX6pfv341aqMyGRkZ6tKli+666y6tX79eDz30kA4fPqwXXnhBkrRp0yZdeOGFOuWUUzRr1izFxMTo22+/1dq1a0NtBINBjRkzRmvWrNHkyZPVu3dvbdy4UQ888IC2bNkSOlc8b948ZWZm6vTTT9fkyZMlSWlpaaF2evfuraFDh+q99947bp83btyokSNHKiUlRbfddpv8fr9mzJihtm3bltv3jjvu0C233KKMjAxlZmbqwIEDevjhh3X22Wfrs88+U1JSkubMmaMXXnhBr776amjI/ZRTTpFUOv8gPj5ef/zjHxUfH693331Xt956q44cOaJ77723Nk+9JGnKlCn64YcftGLFCs2bN6/W7QGNQaTqaHZ2tkpKSrRv3z7NmTNHR44c0fDhw2vUJ+poOOpohJl6NnfuXCPJfPzxx6FtM2bMMJLMgQMHKrzP4cOHjSQzbty40LaVK1caSWbGjBk16sc111xjXC6X2bJlS43uv3z5cuNyuYzL5TJnnnmmmT59ulm2bJnx+Xxh++3YscNIMnPnzi3XxrH9L3sexowZE7bf1KlTjSSzYcMGY4wxDzzwwHGfL2OMmTdvnrFt26xevTps+xNPPGEkmbVr14a2xcXFmYkTJ1bYjiQzdOjQSo9TZuzYscbj8Zhdu3aFtn311VfG5XKZo3/Ndu7caVwul7njjjvC7r9x40YTFRUVtr2y34uCgoJyx58yZYqJjY01RUVFoW2dO3eu8HENHTo07DFV9DO69tprTQO8PIAqacg62rNnTyPJSDLx8fHmr3/9qwkEAjV6HNTRcNTRyGr0p6gkKT4+XpKUm5sb2jZs2DAZY2o0evPyyy/rmWee0Z/+9CedfPLJNerTiBEjtG7dOo0ZM0YbNmzQ3//+d40aNUrt27fXG2+8UaM2y1x77bVht8sm2i1evFjSz+dSX3/9dQWDwQrbWLBggXr37q1evXopKysr9HXuuedKklauXFmlvhhjTviuIxAIaNmyZRo7dqw6deoU2t67d2+NGjUqbN9FixYpGAwqIyMjrF/t2rXTySefXKV+HT0pPDc3V1lZWRoyZEhouB1AeZGqo3PnztXSpUv12GOPqXfv3iosLCx3WqOqqKM/o45GXpM4RZWXlydJSkhIqHVbq1ev1jXXXKNRo0bpjjvuqFVbAwcO1KJFi+Tz+bRhwwa9+uqreuCBBzR+/Hh9/vnn6tOnT43aPTZ0paWlybZt7dy5U5J02WWX6emnn1ZmZqb+53/+R8OHD9cll1yi8ePHy7ZLM+vWrVv19ddfKyUlpcJjRHIC34EDB1RYWFhhWOzZs2eooJT1yxhTabCMjo4+4fE2bdqkv/71r3r33Xd15MiRsO/l5ORUs/dA8xCpOnrmmWeG/j9hwgT17t1bknTffffVqD3qaCnqaOQ1iYBTdp2F2i5x27Bhg8aMGaN+/fpp4cKFioqKzMN3u90aOHCgBg4cqB49euiqq67SggULNGPGjEonVVXnHc+xbXi9Xq1atUorV67UW2+9paVLl2r+/Pk699xztXz5crlcLgWDQfXv31+zZ8+usM2aLhWtrWAwKMuytGTJErlcrnLfL3uXWZns7GwNHTpUiYmJmjVrltLS0uTxePTpp5/qhhtuCHsndrznvqJjA04WqTp6tOTkZJ177rl66aWXahxwylBHq446WjVNIuCUTVA6dpiuOrZt26bzzz9fbdq00eLFi0/4C1BTp512miRp7969kkoLgKRyqwx27dpVaRtbt25V165dQ7e//fZbBYNBdenSJbTNtm0NHz5cw4cP1+zZs3XnnXfq5ptv1sqVK3XeeecpLS1NGzZs0PDhw084c722M9tTUlLk9Xq1devWct/75ptvwm6npaXJGKOuXbuqR48e1T7We++9p4MHD2rRokU6++yzQ9t37NhRbt/k5OQKV3fs2rVL3bp1O+5xWDUFp4lEHa1IYWFhxN/xU0fDUUdrptHPwXn55Zf19NNP68wzzwybqV+d5Y379u3TyJEjZdu2li1bVulwY3WsXLlSxphy28uGEXv27ClJSkxMVOvWrbVq1aqw/R577LFK2y5bWlem7IqeF1xwgSTp0KFD5e4zYMAASQotXczIyNCePXv01FNPldu3sLBQ+fn5odtxcXGVLvOsyvJGl8ulUaNG6bXXXgvb9+uvv9ayZcvC9r3kkkvkcrk0c+bMcs+fMUYHDx484bHK9i3j8/kqfD7T0tK0fv16+Xy+0LY333wzbBllZeLi4iSVL6hAUxSJOlrR6ZidO3fqnXfeCQWS6qKO/ow6GnmNagRn4cKFio+Pl8/nC12Bc+3atTr11FO1YMGCsH2rs7zx/PPP1/bt2zV9+nStWbNGa9asCX2vbdu2oSWTUuna/eeff147duwIS/rHmjZtmgoKCjRu3Dj16tVLPp9PH3zwgebPn68uXbroqquuCu2bmZmpu+++W5mZmTrttNO0atUqbdmypdK2d+zYoTFjxuj888/XunXr9OKLL+ryyy8PXSl01qxZWrVqlUaPHq3OnTtr//79euyxx9ShQwcNHjxYUumS0FdeeUW/+93vtHLlSg0aNEiBQECbN2/WK6+8omXLloWKUnp6ut5++23Nnj1bqamp6tq1q37xi19IqvryxpkzZ2rp0qUaMmSIpk6dKr/fr4cfflh9+/bVF198EdovLS1Nf/vb33TjjTdq586dGjt2rBISErRjxw69+uqrmjx5sv785z9XepyzzjpLycnJmjhxoq677jpZlqV58+ZVWCQzMzO1cOFCnX/++crIyNC2bdv04osvhi3frEx6erok6brrrtOoUaPkcrk0YcKEE94PaGh1VUf79++v4cOHa8CAAUpOTtbWrVv1zDPPqKSkRHfffXfYvtRR6mjZcyI1YB2t72Vbx1veWPbl8XhMhw4dzIUXXmieffbZsCVrZaqzvPHoto/9Onbp3qWXXmq8Xq85fPjwcdtcsmSJufrqq02vXr1MfHy8cbvdpnv37mbatGnmxx9/DNu3oKDAXHPNNaZFixYmISHBZGRkmP3791e6vPGrr74y48ePNwkJCSY5Odn84Q9/MIWFhaH93nnnHXPxxReb1NRU43a7TWpqqvnVr35Vbsm7z+cz99xzj+nbt6+JiYkxycnJJj093cycOdPk5OSE9tu8ebM5++yzjdfrNZLClgRW9BxV5v333zfp6enG7Xabbt26mSeeeCL0mI71r3/9ywwePNjExcWZuLg406tXL3Pttdeab775ptzzcezyxrVr15ozzjjDeL1ek5qaGlpaKsmsXLkybN/777/ftG/f3sTExJhBgwaZTz75pErLG/1+v5k2bZpJSUkxlmWxZByNSkPU0RkzZpjTTjvNJCcnm6ioKJOammomTJhgvvjii3L7Ukepo8Y0fB21jKkgsjVjbdu21ZVXXhmRCx0BQHNEHUVjQMA5yqZNm3TmmWdq+/btNb7CMQA0Z9RRNBYEHAAA4DiNfhUVAABAdRFwAACA4xBwAACA4xBwAACA4xBwAACA41T5SsYN/ZkSgBOwaLF5o44CtVfVOsoIDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcBwCDgAAcJyohu4AnMFlSd7oumm7yC/5g3XTNgA0lGhbimmgv8I+v+RzeF0l4CAiBrSz9dhoT50MCU5/u1grdwbqoGUAaDjjekfpL2e6G+TYT35aoqc+LWmQY9cXAg4iIjHGUvpJtrYdMtp8MLJvCwr9JqLtAUBjkO8z+iGvfutbarylU9vZOineqtfjNgQCDiLqvnU+Pb+hRCaCr1lOTwFwoiXfBrR8W2G9HnNAO1srroit12M2FAIOIiqrICgfZ5MA4ISCpvSrvrgs6Uix0cc/BPVDrvNHxgk4AAA4XIdES7cMcWvF9oAumV+gIn9D96jusUwcAACHa+W1NHFAtFrFWsr1SSXN4NQ/AQcAADgOAQcAAIezJB0qlIqb0apU5uAAAOBQ7RMsXdwzSlsOBXXmM/k6WNh8Ag4jOIgYX0AKNJ/XDgA0ej1a2XroghgluKVdOUZ5vobuUf1hBAe10iHR0hntXTpQENTpTxfou5xmMHMNANDoMYKDWjmro0v/O96joJG+3B/UkeKG7hEAAIzgAADgOG1iLfVtYytojM55vkBfHWh+o+uM4AAA4DDDurq04gqvgkZa/V1QB+v3EyEaBUZwUCei7dLJbVFVjND+oLTlYLBZXHwKAFD3CDioE6kJlt65MlZJnqrtn1UgnfFMvr4/wjIsAEDtEXAQEae3j9L3R0pUFk9SYi39dWWxXFbV7h8ISu3irLARn725RsV8cCcAVFliTOnHMhwuNJr6VrG+PdR83zRaxpgqPXrLquJfKjQrGX2j9PIlHhX5FfYp4tPfLtbTn5ZUuR1L0u9Oi9Yd58ZIKv2E3Yv+WaB13zvrnFUVX25wKOoo6trk/xet24bFaOSLBfpyv7PqZ5mq1lFGcBARa3cHtP77nxPO53urN/RiJH26N6CHPyq9CpUxUrRLah1r6UixCQtPAICKxURJSZ7SGtrcEXAQEfd/4NOK7T+nkJq8tj7cE9RHe36+zObQLi59eE2spi4u1LJtJBwAQNURcFAr2w8F9dSnJdqeHaxRqDnW0W3szQ3q7e0lsi0pwS0V+cUqKwCogNsleaOkH/ONnv2sRIeb0WdOVYY5OKg12yqdM1NXbZ/a1tYj/+nR/R/4tGizv24OVE+Yg9O8UUdRFyxJvz8tWpf2jtbkNwu1I9vUWU1uDKpaR7nQH2qtLl9IQVP6lX6SrZQ4/jgAQEVSE231b2vLF6jbmtyUEHAAAIDjEHDQ6OWXGH20J6j8EqMou3Q4FgBQWg+jXdKhAqNP9wZYcXoU5uCg0bMleaOlrsm2pp4WraXf+vXGlqb5KmYOTvNGHUUkWZImDYhSvxSXHv3Ep315RoUlNVvF2pQwBweOEZSUXyLl+Ywm9IvWgHauhu4SADQ425IGdYrShT2ilFNkVNAMwk11EHAAAIDjEHAAAGiCLEsqKjE6UuzsZeE1xYX+AABoQmxL+s/uLrWOtfT4JyW6c41POcUN3avGh4ADAEATEmVLvzvNrVZeSwu+KlB+1T/XuFnhFBUAAHAcAg4AAE0QH813fJyiAgCgCXBZ0oB2thJiLP19bbGOFEuFTfvj+eoUAQcAgCbAGy09Mdqj73ONMhYUqoQhnOMi4KBGWsdKJ8Xb2nIwqOI6vqiwJ0rq0cpWvFu6+vVCbT7IqxpAM2WVrqLCiTEHBzWS0Sdaq6+KVVrLuv8V6tHK1pqrYtUvxaXXvglocxYXfAAAHB8jOKgRly25XfXzwZeWfjoW71oANEO2JaUmWEr2WHroQ58O5BsFeJ93QgQcAAAasZYe6c1fefXezoD+vKJYfs7SVwmnqFAjH+0J6O41PsVFl87Hia6D36RoW2oda8kbJd25xqf/+6FpfoI4ANSGZVlq4bEVEyU+kqEaCDiokQ/3BHX7Kp86trC17pp4DWgX+V+l9FRbH14TpzZxlm5/36dP9vK2BQBQNZyiQo0ZSbuyg3p3e4m80VK8++fvFftV7SWM0bYUc9RvpMdl6e3tJdp9xIg3LQCaG0tSbLSU5JEWflWiz/cFZCiGVWYZU7Wny2KGJyphW9LZnV2657yY0JDgXWt8WrS5elegyugbpb+cWZqSgpL+srxIa3YHHTUkW8WXGxyKOorqSI23NO8Sr974pkSPfVwif1C82VPV6ygjOKi1oJGyCox+yDWh6zPERJVev6aq4cRlSW5b+iHPhNo8WMj5ZgDNlztK+o92tt7faXFRvxog4CAiNu0PKmNBYeh295a2XhjrUWx01d6x5vuMbllZrPmbfh71YaUAAKCmCDiICKPwOTfZRUYtPLYS3ZXeJUyULeUUV3/eDgA4VbFf+viHoH7IZSi7JpiDgzphqfRzU6r6W2MkFZY4//wyc3CaN+ooqqOsjpYEePN3tKrWUQIOUI8IOM0bdRSovarWUa6DAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHIeAAwAAHCeqoTtQb2yXLFe0TEmxJNPQvQGARseKckvWid/3UkfRFDSbgBPb4ywlpF+kg2/Nlj97X0N3BwAaFSsqRi1HXavoVh2Ou58J+KmjaBKcH3AsW5YrStGtOsrdNk2Kcjd0jwCgcbCjJKv0v1ZMrIK+QvnzDh3/PsGALHes5Drqz0cgIEZ00Ng4PuDE9jlHcX3Olrt154buCgA0GnZsklqeN0WW2yNJKt69SdnvzZUJ+I9/R0vydk1X0tlXhDblrHtFvj1f12V3gWpzfMCJbpmq2G7pkqRgSXED9wYAGgHLlu2Jk+1NkBXt+WmjqfLcGlNSLNuToKiElrLjWylvw/I67S5QE44POACAn7kSWqvFWRNUuPMzHXj1DsmUBhoT9Kuqp5mKvvtCxfP/Km+Ps9TqP6+vw94CNefsZeKWLVNSJN+hHxQszm/o3gBAg7M9CYrrN1wub6KMr1CmpEimpEg60ampo5mgTEmR/If2yJ/9o2SCCk3mARoJhwYcS94eZyrxF5eqYMs6/fjSdOV9saKhOwUAjuL78Vv9OO9PCuRnq8XgX8t9Uo+G7hIQ4syAY9vydB6g2B5nKViQrWBBdum7DABo7iwpWJQr4/fVvq1gQMGiXAVLihR/6khFt+pY+zaBCHHWHBzLlrdbumxvC+V+9paOrJvPqSkAkOSKbyXvyWfIf2iP9s37k4KFuRFruy7aBGrLWSM4tktxp45S/CnnKZD9owJ5h0IT6ACgOYtqmark8ybLcnsVOHKgdN5NpAT9kW8TqCVnjOBYtjydTpEV5VbO2n/KFOdHZvgVAAA0SY4YwbGi3UoaOlFxp45UyYFdP11CnJEbAACaK0cEHAAAgKMRcAAAgOM07Tk4lq2olu1lu73KWftPBQpzpGCgoXsFAI2GFROnqMQ2kjHKeu1uFe/9pqG7BNSLJj2CY3vilXLJLYrrd64Kt38i3w/fiLk3APAzT5cBanfFfZIxKty6TsETfVo44BBNcwTHsuVKaCXbk6Aj61+R//Deny4VDgA4mmXZkqsqpd6SKz5Zsivb1yiQe5BaiyajSQYcV0IrtfvVXcrdsFxH1i8QozYAUDtWdIxaXfhnudt0rfD7gcJc7f/njQrkHaznngE10yQDjmTJiomTXC4RbgCgBixLdky8ZJV+SKbtiVfhtx+qePemCnc3wRLZ3kSZQEloW7A4n3mPaLSaaMABANSGK76V2lx6q6yYWElSweY1yl79ohSs5FPFLUvxp4xS4hnjJUnGBHXwjfvk27elvroMVEvTCjiWJSsqRrbbq/xNK1Wyb1tD9wgAmpayOhoTp6LdX8qKipYk+Q7sPP5ojDEqOfyDinZ+9nNT0dGyoj2lV45nbg4amSYVcKKSTlKr0f+t3M+W6PC7T/OCAoBqKldHy07zG+lEp/yLv/tCxbs3/nTLUkzHvmrzqzuU/d5zKv5u43HvC9S3JhJwLMnlkh0Tq0DeYQULcgg3AFANdlyS5IqqfR0NfYCxUbDgiAJ5h2VFeSRXdOnpLT7gGI1Ek7gOTnRKZ6WMu0l2XLKy/v13Fe38tKG7BABNStLgXytl3M0RraMlWd8p67W7Fcw/rNYX3yBP1/QI9BSIjMYfcCxbtjdBtidBxlckBXiHAADVYdsuuVt3UmzaQFl2VATrqJGCfplgQN6u6YpKTIlAm0BkNOpTVO52PRQ/4Hzlb3pH++ffUjqRDQAA4AQa9QhOVFJbxfUfLhnJlBQx7wYAaiAYDMhfkCPfoR9Ka2mEmUCJ/Nl7ZfwloevqAA2tUY/gAAAiI3vVCyr85gMFfYURb9t/+Af9+NJ0uRJT1OKsCSr6fpOKd30R8eMA1dG4A47tCv+3Vm2VDlZZkWgLAJoY/8HvFSzKrZvGTVDBojxZ7lglpI+RkQg4aHCNMuBEp3SRp2N/ebsPlCQl/Md/KrpVxxq3Z7lciu05SLJdSjjtYvn2bj3hfUqyvjvqeg8AAKApaZQBJ6ZDHyUNm1R6I+CXN22gvGkDa9eo7ZIsS3F9hiqu15AT7p73xXIVf/8lK7YAAGiCGmXAKfh6daUf+FYjtkuJA8cqttcQZb/3rIqqMHQaLMol3AAA0EQ1yoATLMqN+Lli396tiu05SIW7Nsif9V1E2wYAAI1Lo14mDgAAUBMEHAAA4DiN8hQVACAyir//Slmv3S1ZlqJTusp/5EeZ4oLIHiTKrejkVFlurw4ueVAlB7+PbPtADTCCAwAOFsg7qMKt6yTLUrsr7pOnY/+IHyM6OVXtfn2v3K07q3DrevkPEXDQ8Ag4ANBcuKIkqy7KvvVT23xMAxoPAg4AAHAcAg4ANAP+nP068sF8mZIi2d4Wkiu69o3aLtneRFnRbuWsf0W+fd/Wvk0gQgg4ANAMBHJ+VM7alxUoyFHb3/xdsT3PqnWb7nbd1e7K2XLFJunI2n/Kt+/EH4MD1BdWUQFAMxIszFHRzs9lAn5Z7ljJBCVJJhiQAiXHv7PtknXUyI/lcqto52fyHzlQl10GaoSAAwDNSCD3oA6//YSiklKVcslfZbs9kqSCLet0ZP1CSZV/RE1sj7OUePo4SZKRUfa7z+rQ8sdDIQloTAg4ANDcGKNgcb6C+YcU9Hl+2haU5fbIVDaKY9mSK0r+vEOhTYGiXMINGi0CDgA0Q8GCbGW9NSd02/YmKHnY1XIltq5wf+MrUvbqF1Xw9eqjGvHXcS+Bmms2ASeQf1i+vVulEl9DdwUAGoejAoopLpCiomV7Eirc1Vi2gsX5hBo0GZYxpvITrkfv2NQv4GS7ZNlRMn6fjneOGahLVXy5waEafR2Ncsuq7EKAxsj4i+u3P0AFqlpHm0/AARoBAk7zRh0Faq+qdZTr4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMch4AAAAMexjDGmoTsBAAAQSYzgAAAAxyHgAAAAxyHgAAAAxyHgAAAAxyHgAAAAxyHgAAAAxyHgAAAAxyHgAAAAxyHgAAAAx/n/+5VmB8PqrzIAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 800x600 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize the coco_instances exported dataset.\n",
    "# In this time, we can see each mask has its class label.\n",
    "imported_dataset = dm.Dataset.import_from(\"ws_test/inst-mask\", \"coco_instances\")\n",
    "viz = dm.Visualizer(imported_dataset, ignored_types=[dm.AnnotationType.bbox])\n",
    "fig = viz.vis_gallery([item_id for item_id in range(4)], dm.DEFAULT_SUBSET_NAME)\n",
    "fig.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "datumaro",
   "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": "3.9.13"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}