00-open-images-data-manipulation.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Open Images data preparation\n",
    "\n",
    "The goal of this notebook is to get the public Open Images data formatted in such a way it can be used with Detectron2.\n",
    "\n",
    "To download the Open Images, we're using the following script:\n",
    "- https://github.com/spmallick/learnopencv/blob/master/downloadOpenImages/downloadOI.py\n",
    "\n",
    "We need to format the images downloaded using the script into Detectron2 style:\n",
    "- https://detectron2.readthedocs.io/tutorials/datasets.html\n",
    "\n",
    "Detectron2 example style:\n",
    "```\n",
    "{'annotations': [{'bbox': [994, 619, 1445, 1166],\n",
    "   'bbox_mode': <BoxMode.XYXY_ABS: 0>,\n",
    "   'category_id': 0,\n",
    "   'iscrowd': 0,\n",
    "   'segmentation': [[1020.5,\n",
    "     963.5,\n",
    "     1000.5,\n",
    "     841.5,\n",
    "\t\t ... #### Truncated ####\n",
    "     989.5,\n",
    "     1020.5,\n",
    "     963.5]]}],\n",
    " 'file_name': 'balloon/train/34020010494_e5cb88e1c4_k.jpg',\n",
    " 'height': 1536,\n",
    " 'image_id': 0,\n",
    " 'width': 2048}\n",
    "```\n",
    "\n",
    "An example notebook using Detectron2 style inputs can be found here: https://colab.research.google.com/drive/16jcaJoc6bCFAQ96jDe2HwtXj7BMD_-m5\n",
    "\n",
    "We'll start by downloading a few sample images, then build a little script which converts the samples to Detectron2 style before ramping it up."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # Install wget\n",
    "# import sys\n",
    "# !conda install --yes --prefix {sys.prefix} -c anaconda wget"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # Download required meta-files\n",
    "# !wget https://storage.googleapis.com/openimages/2018_04/class-descriptions-boxable.csv\n",
    " \n",
    "# !wget https://storage.googleapis.com/openimages/2018_04/train/train-annotations-bbox.csv\n",
    " \n",
    "# !wget https://storage.googleapis.com/openimages/2018_04/validation/validation-annotations-bbox.csv\n",
    " \n",
    "# !wget https://storage.googleapis.com/openimages/2018_04/test/test-annotations-bbox.csv "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Classes of amenities Airbnb mostly cares about\n",
    "subset = [\"Toilet\",\n",
    "          \"Swimming_pool\",\n",
    "          \"Bed\",\n",
    "          \"Biliard_table\",\n",
    "          \"Sink\",\n",
    "          \"Fountain\",\n",
    "          \"Oven\",\n",
    "          \"Ceiling_fan\",\n",
    "          \"Television\",\n",
    "          \"Microwave_oven\",\n",
    "          \"Gas_stove\",\n",
    "          \"Refrigerator\",\n",
    "          \"Kitchen_&_dining_room_table\",\n",
    "          \"Washing_machine\",\n",
    "          \"Bathtub\",\n",
    "          \"Stairs\",\n",
    "          \"Fireplace\",\n",
    "          \"Pillow\",\n",
    "          \"Mirror\",\n",
    "          \"Shower\",\n",
    "          \"Couch\",\n",
    "          \"Countertop\",\n",
    "          \"Coffeemaker\",\n",
    "          \"Dishwasher\",\n",
    "          \"Sofa_bed\",\n",
    "          \"Tree_house\",\n",
    "          \"Towel\",\n",
    "          \"Porch\",\n",
    "          \"Wine_rack\",\n",
    "          \"Jacuzzi\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Install tqdm & awscli\n",
    "#!conda install --yes --prefix {sys.prefix} -c conda-forge tqdm awscli"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Class 0 : Bathtub\n",
      "Class 1 : Toilet\n",
      "Annotation Count : 43\n",
      "Number of images to be downloaded : 39\n",
      "100%|███████████████████████████████████████████| 39/39 [00:50<00:00,  1.29s/it]\n"
     ]
    }
   ],
   "source": [
    "# Download bathtub and toilet images\n",
    "!python3 downloadOI.py --classes 'Bathtub,Toilet' --mode validation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Converting images/labels to Detectron2 style\n",
    "Now I've got images ready, my goal is to now convert their current style them into Detectron2 style (see below).\n",
    "\n",
    "Current style:\n",
    "- Filepath: `validation/Toilet/0d0719cfd8e417b7.jpg`\n",
    "- Label filename: `0d0719cfd8e417b7.txt`\n",
    "- Image filename: `0d0719cfd8e417b7.jpg`\n",
    "- Contents of label file: `Toilet,0.093371,0.986232,0.189984,0.965806`\n",
    "\n",
    "I'll need a way to:\n",
    "1. Traverse different file paths\n",
    "2. Gather filenames and explore their text contents\n",
    "3. Seek duplicates and make sure they contain multiple labels\n",
    "\n",
    "What I should end up with is:\n",
    "- A single file of images (can convert this to train/test later)\n",
    "- A dictionary of all of the different parameters for each image path\n",
    "    - Perhaps I could do this with a Pandas dataframe? So it's visual? Then convert it to JSON maybe?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'BoxMode' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-14-0012c7986773>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      1\u001b[0m detectron2_style = {\n\u001b[1;32m      2\u001b[0m    'annotations': [{'bbox': [994, 619, 1445, 1166],\n\u001b[0;32m----> 3\u001b[0;31m    \u001b[0;34m'bbox_mode'\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mBoxMode\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mXYXY_ABS\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      4\u001b[0m    \u001b[0;34m'category_id'\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m    \u001b[0;34m'iscrowd'\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'BoxMode' is not defined"
     ]
    }
   ],
   "source": [
    "detectron2_style = {\n",
    "   'annotations': [{'bbox': [994, 619, 1445, 1166],\n",
    "   'bbox_mode': BoxMode.XYXY_ABS,\n",
    "   'category_id': 0,\n",
    "   'iscrowd': 0,\n",
    "   'segmentation': [[1020.5,\n",
    "     963.5,\n",
    "     1000.5,\n",
    "     841.5,\n",
    "     #### Truncated ####\n",
    "     989.5,\n",
    "     1020.5,\n",
    "     963.5]]}],\n",
    " 'file_name': 'balloon/train/34020010494_e5cb88e1c4_k.jpg',\n",
    " 'height': 1536,\n",
    " 'image_id': 0,\n",
    " 'width': 2048\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: []\n",
       "Index: []"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "df = pd.DataFrame()\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(50, 25)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import os\n",
    "file_ids = [os.path.splitext(file)[0] for file in os.listdir(\"validation/Toilet\")]\n",
    "len(file_ids), len(set(file_ids))       "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "unique_ids = list(set(file_ids))\n",
    "#unique_ids"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "df[\"image_id\"] = unique_ids"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 569 µs, sys: 1.07 ms, total: 1.64 ms\n",
      "Wall time: 1.97 ms\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(['validation/Toilet/d60e1c25e87d8b45.jpg',\n",
       "  'validation/Toilet/17f698ec871569ca.jpg',\n",
       "  'validation/Toilet/7c2825a3d8e0ed29.jpg',\n",
       "  'validation/Toilet/40a69f79da2aeaaa.jpg',\n",
       "  'validation/Toilet/b2dc8e2437e8803f.jpg',\n",
       "  'validation/Toilet/e6e05b56799fefba.jpg',\n",
       "  'validation/Toilet/c1eefc708000b69f.jpg',\n",
       "  'validation/Toilet/b3a65783709539de.jpg',\n",
       "  'validation/Toilet/0d0719cfd8e417b7.jpg',\n",
       "  'validation/Toilet/3454c87a14067798.jpg',\n",
       "  'validation/Toilet/e108d68162990dfb.jpg',\n",
       "  'validation/Toilet/2f0a18a409d5a769.jpg',\n",
       "  'validation/Toilet/217b95a71cb220f8.jpg',\n",
       "  'validation/Toilet/3a511ef0f68cc438.jpg',\n",
       "  'validation/Toilet/be84a3da54658167.jpg',\n",
       "  'validation/Toilet/36d8c654fba5f337.jpg',\n",
       "  'validation/Toilet/47ab0e73c33bdefd.jpg',\n",
       "  'validation/Toilet/c7312a8b436f5e90.jpg',\n",
       "  'validation/Toilet/b821cc12eede040e.jpg',\n",
       "  'validation/Toilet/85ccc93d1f941931.jpg',\n",
       "  'validation/Toilet/eaa75967cbc70ac1.jpg',\n",
       "  'validation/Toilet/2ff2bb609a057f7a.jpg',\n",
       "  'validation/Toilet/539e0871494dea5d.jpg',\n",
       "  'validation/Toilet/42298c9659ce6603.jpg',\n",
       "  'validation/Toilet/f90eac5444f62b19.jpg',\n",
       "  'validation/Bathtub/2f2039140c8b1f2b.jpg',\n",
       "  'validation/Bathtub/784bba5f12ee7c35.jpg',\n",
       "  'validation/Bathtub/837b3d11ff02f116.jpg',\n",
       "  'validation/Bathtub/2f0a18a409d5a769.jpg',\n",
       "  'validation/Bathtub/789cc0283c0e18af.jpg',\n",
       "  'validation/Bathtub/eccb080e57b2aac5.jpg',\n",
       "  'validation/Bathtub/11dcac4ca5923a58.jpg',\n",
       "  'validation/Bathtub/15348e4f2c7ebe0f.jpg',\n",
       "  'validation/Bathtub/822f20d881eebeb9.jpg',\n",
       "  'validation/Bathtub/ddd3f96ff7f0bc78.jpg',\n",
       "  'validation/Bathtub/d7d469cb4c7e8cd2.jpg',\n",
       "  'validation/Bathtub/8e562fee6ff208f7.jpg',\n",
       "  'validation/Bathtub/cd727c5e0d2cfc1b.jpg',\n",
       "  'validation/Bathtub/3f66fdf9688514d4.jpg'],\n",
       " ['validation/Toilet/e108d68162990dfb.txt',\n",
       "  'validation/Toilet/217b95a71cb220f8.txt',\n",
       "  'validation/Toilet/2f0a18a409d5a769.txt',\n",
       "  'validation/Toilet/36d8c654fba5f337.txt',\n",
       "  'validation/Toilet/47ab0e73c33bdefd.txt',\n",
       "  'validation/Toilet/c7312a8b436f5e90.txt',\n",
       "  'validation/Toilet/be84a3da54658167.txt',\n",
       "  'validation/Toilet/3a511ef0f68cc438.txt',\n",
       "  'validation/Toilet/7c2825a3d8e0ed29.txt',\n",
       "  'validation/Toilet/d60e1c25e87d8b45.txt',\n",
       "  'validation/Toilet/17f698ec871569ca.txt',\n",
       "  'validation/Toilet/b2dc8e2437e8803f.txt',\n",
       "  'validation/Toilet/40a69f79da2aeaaa.txt',\n",
       "  'validation/Toilet/b3a65783709539de.txt',\n",
       "  'validation/Toilet/e6e05b56799fefba.txt',\n",
       "  'validation/Toilet/c1eefc708000b69f.txt',\n",
       "  'validation/Toilet/0d0719cfd8e417b7.txt',\n",
       "  'validation/Toilet/3454c87a14067798.txt',\n",
       "  'validation/Toilet/539e0871494dea5d.txt',\n",
       "  'validation/Toilet/42298c9659ce6603.txt',\n",
       "  'validation/Toilet/f90eac5444f62b19.txt',\n",
       "  'validation/Toilet/85ccc93d1f941931.txt',\n",
       "  'validation/Toilet/b821cc12eede040e.txt',\n",
       "  'validation/Toilet/2ff2bb609a057f7a.txt',\n",
       "  'validation/Toilet/eaa75967cbc70ac1.txt',\n",
       "  'validation/Bathtub/837b3d11ff02f116.txt',\n",
       "  'validation/Bathtub/2f0a18a409d5a769.txt',\n",
       "  'validation/Bathtub/eccb080e57b2aac5.txt',\n",
       "  'validation/Bathtub/789cc0283c0e18af.txt',\n",
       "  'validation/Bathtub/2f2039140c8b1f2b.txt',\n",
       "  'validation/Bathtub/784bba5f12ee7c35.txt',\n",
       "  'validation/Bathtub/d7d469cb4c7e8cd2.txt',\n",
       "  'validation/Bathtub/8e562fee6ff208f7.txt',\n",
       "  'validation/Bathtub/3f66fdf9688514d4.txt',\n",
       "  'validation/Bathtub/cd727c5e0d2cfc1b.txt',\n",
       "  'validation/Bathtub/11dcac4ca5923a58.txt',\n",
       "  'validation/Bathtub/822f20d881eebeb9.txt',\n",
       "  'validation/Bathtub/15348e4f2c7ebe0f.txt',\n",
       "  'validation/Bathtub/ddd3f96ff7f0bc78.txt'])"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "img_files = []\n",
    "label_files = []\n",
    "for path, dirnames, filenames in os.walk(\"validation\"):\n",
    "    for file in filenames:\n",
    "        if \".jpg\" in file:\n",
    "            #print(os.path.join(path, file))\n",
    "            img_files.append(os.path.join(path, file))\n",
    "        else:\n",
    "            label_files.append(os.path.join(path, file))\n",
    "            #print(os.path.join(path, file))\n",
    "        #img_files = set(img_files.append(os.path.join(path, file)))\n",
    "img_files, label_files"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['validation', 'Toilet', 'd60e1c25e87d8b45']"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "os.path.splitext(img_files[0])[0].split(\"/\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 127 µs, sys: 1 µs, total: 128 µs\n",
      "Wall time: 131 µs\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[{'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': 'd60e1c25e87d8b45',\n",
       "  'file_name': 'validation/Toilet/d60e1c25e87d8b45.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '17f698ec871569ca',\n",
       "  'file_name': 'validation/Toilet/17f698ec871569ca.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '7c2825a3d8e0ed29',\n",
       "  'file_name': 'validation/Toilet/7c2825a3d8e0ed29.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '40a69f79da2aeaaa',\n",
       "  'file_name': 'validation/Toilet/40a69f79da2aeaaa.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': 'b2dc8e2437e8803f',\n",
       "  'file_name': 'validation/Toilet/b2dc8e2437e8803f.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': 'e6e05b56799fefba',\n",
       "  'file_name': 'validation/Toilet/e6e05b56799fefba.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': 'c1eefc708000b69f',\n",
       "  'file_name': 'validation/Toilet/c1eefc708000b69f.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': 'b3a65783709539de',\n",
       "  'file_name': 'validation/Toilet/b3a65783709539de.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '0d0719cfd8e417b7',\n",
       "  'file_name': 'validation/Toilet/0d0719cfd8e417b7.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '3454c87a14067798',\n",
       "  'file_name': 'validation/Toilet/3454c87a14067798.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': 'e108d68162990dfb',\n",
       "  'file_name': 'validation/Toilet/e108d68162990dfb.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '2f0a18a409d5a769',\n",
       "  'file_name': 'validation/Toilet/2f0a18a409d5a769.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '217b95a71cb220f8',\n",
       "  'file_name': 'validation/Toilet/217b95a71cb220f8.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '3a511ef0f68cc438',\n",
       "  'file_name': 'validation/Toilet/3a511ef0f68cc438.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': 'be84a3da54658167',\n",
       "  'file_name': 'validation/Toilet/be84a3da54658167.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '36d8c654fba5f337',\n",
       "  'file_name': 'validation/Toilet/36d8c654fba5f337.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '47ab0e73c33bdefd',\n",
       "  'file_name': 'validation/Toilet/47ab0e73c33bdefd.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': 'c7312a8b436f5e90',\n",
       "  'file_name': 'validation/Toilet/c7312a8b436f5e90.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': 'b821cc12eede040e',\n",
       "  'file_name': 'validation/Toilet/b821cc12eede040e.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '85ccc93d1f941931',\n",
       "  'file_name': 'validation/Toilet/85ccc93d1f941931.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': 'eaa75967cbc70ac1',\n",
       "  'file_name': 'validation/Toilet/eaa75967cbc70ac1.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '2ff2bb609a057f7a',\n",
       "  'file_name': 'validation/Toilet/2ff2bb609a057f7a.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '539e0871494dea5d',\n",
       "  'file_name': 'validation/Toilet/539e0871494dea5d.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '42298c9659ce6603',\n",
       "  'file_name': 'validation/Toilet/42298c9659ce6603.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': 'f90eac5444f62b19',\n",
       "  'file_name': 'validation/Toilet/f90eac5444f62b19.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': '2f2039140c8b1f2b',\n",
       "  'file_name': 'validation/Bathtub/2f2039140c8b1f2b.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': '784bba5f12ee7c35',\n",
       "  'file_name': 'validation/Bathtub/784bba5f12ee7c35.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': '837b3d11ff02f116',\n",
       "  'file_name': 'validation/Bathtub/837b3d11ff02f116.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': '2f0a18a409d5a769',\n",
       "  'file_name': 'validation/Bathtub/2f0a18a409d5a769.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': '789cc0283c0e18af',\n",
       "  'file_name': 'validation/Bathtub/789cc0283c0e18af.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': 'eccb080e57b2aac5',\n",
       "  'file_name': 'validation/Bathtub/eccb080e57b2aac5.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': '11dcac4ca5923a58',\n",
       "  'file_name': 'validation/Bathtub/11dcac4ca5923a58.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': '15348e4f2c7ebe0f',\n",
       "  'file_name': 'validation/Bathtub/15348e4f2c7ebe0f.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': '822f20d881eebeb9',\n",
       "  'file_name': 'validation/Bathtub/822f20d881eebeb9.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': 'ddd3f96ff7f0bc78',\n",
       "  'file_name': 'validation/Bathtub/ddd3f96ff7f0bc78.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': 'd7d469cb4c7e8cd2',\n",
       "  'file_name': 'validation/Bathtub/d7d469cb4c7e8cd2.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': '8e562fee6ff208f7',\n",
       "  'file_name': 'validation/Bathtub/8e562fee6ff208f7.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': 'cd727c5e0d2cfc1b',\n",
       "  'file_name': 'validation/Bathtub/cd727c5e0d2cfc1b.jpg'},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': '3f66fdf9688514d4',\n",
       "  'file_name': 'validation/Bathtub/3f66fdf9688514d4.jpg'}]"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "img_details = []\n",
    "for img_path in img_files:\n",
    "    dataset, label, img_id = os.path.splitext(img_path)[0].split(\"/\")\n",
    "    img_detail = {\n",
    "        \"dataset\": dataset,\n",
    "        \"label\": label,\n",
    "        \"id\": img_id,\n",
    "        \"file_name\": img_path\n",
    "    }\n",
    "    img_details.append(img_detail)\n",
    "img_details"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[{'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': 'e108d68162990dfb',\n",
       "  'bbox': ['0.360770', '0.656517', '0.410382', '0.886037']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '217b95a71cb220f8',\n",
       "  'bbox': ['0.438669', '0.602654', '0.449081', '0.532643']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '2f0a18a409d5a769',\n",
       "  'bbox': ['0.000008', '0.546224', '0.349488', '0.999980']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '36d8c654fba5f337',\n",
       "  'bbox': ['0.063034', '0.805088', '0.000000', '0.697051']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '47ab0e73c33bdefd',\n",
       "  'bbox': ['0.197381', '0.922556', '0.468605', '0.829202']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': 'c7312a8b436f5e90',\n",
       "  'bbox': ['0.483064', '0.583230', '0.148464', '0.254009Toilet']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': 'be84a3da54658167',\n",
       "  'bbox': ['0.246185', '0.776473', '0.728370', '1.000000']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '3a511ef0f68cc438',\n",
       "  'bbox': ['0.124107', '0.338616', '0.755953', '0.978802']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '7c2825a3d8e0ed29',\n",
       "  'bbox': ['0.207057', '0.415161', '0.809775', '0.999841']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': 'd60e1c25e87d8b45',\n",
       "  'bbox': ['0.405653', '1.000000', '0.146613', '0.626838']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '17f698ec871569ca',\n",
       "  'bbox': ['0.160562', '0.475538', '0.181890', '0.847251Toilet']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': 'b2dc8e2437e8803f',\n",
       "  'bbox': ['0.337954', '0.695044', '0.258797', '0.857483']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '40a69f79da2aeaaa',\n",
       "  'bbox': ['0.014170', '0.473459', '0.348211', '0.999983']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': 'b3a65783709539de',\n",
       "  'bbox': ['0.184904', '0.846895', '0.060172', '0.961222']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': 'e6e05b56799fefba',\n",
       "  'bbox': ['0.059556', '0.956108', '0.000000', '0.507183']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': 'c1eefc708000b69f',\n",
       "  'bbox': ['0.201985', '0.772443', '0.097975', '0.983168']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '0d0719cfd8e417b7',\n",
       "  'bbox': ['0.093371', '0.986232', '0.189984', '0.965806']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '3454c87a14067798',\n",
       "  'bbox': ['0.077027', '0.306008', '0.811938', '1.000000Toilet']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '539e0871494dea5d',\n",
       "  'bbox': ['0.201278', '0.815655', '0.000000', '1.000000']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '42298c9659ce6603',\n",
       "  'bbox': ['0.301737', '0.792065', '0.190453', '0.838540']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': 'f90eac5444f62b19',\n",
       "  'bbox': ['0.405385', '0.896555', '0.624074', '1.000000']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '85ccc93d1f941931',\n",
       "  'bbox': ['0.369258', '0.746107', '0.152027', '1.000000']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': 'b821cc12eede040e',\n",
       "  'bbox': ['0.520803', '0.940757', '0.603367', '0.988894']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': '2ff2bb609a057f7a',\n",
       "  'bbox': ['0.182173', '0.798462', '0.273742', '0.942715']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Toilet',\n",
       "  'id': 'eaa75967cbc70ac1',\n",
       "  'bbox': ['0.095922', '0.870645', '0.124023', '0.570180']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': '837b3d11ff02f116',\n",
       "  'bbox': ['0.000000', '0.669035', '0.372941', '1.000000']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': '2f0a18a409d5a769',\n",
       "  'bbox': ['0.157351', '1.000000', '0.328732', '0.927573']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': 'eccb080e57b2aac5',\n",
       "  'bbox': ['0.303341', '0.905829', '0.470852', '0.999948']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': '789cc0283c0e18af',\n",
       "  'bbox': ['0.039686', '0.948718', '0.119581', '1.000000']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': '2f2039140c8b1f2b',\n",
       "  'bbox': ['0.000000', '1.000000', '0.000000', '1.000000']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': '784bba5f12ee7c35',\n",
       "  'bbox': ['0.000000', '1.000000', '0.000000', '1.000000']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': 'd7d469cb4c7e8cd2',\n",
       "  'bbox': ['0.000081', '0.999159', '0.435396', '1.000000']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': '8e562fee6ff208f7',\n",
       "  'bbox': ['0.013755', '0.513952', '0.476118', '0.999980Bathtub']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': '3f66fdf9688514d4',\n",
       "  'bbox': ['0.050368', '0.933495', '0.250992', '1.000000']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': 'cd727c5e0d2cfc1b',\n",
       "  'bbox': ['0.048442', '0.980496', '0.427911', '0.806723']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': '11dcac4ca5923a58',\n",
       "  'bbox': ['0.360645', '0.999988', '0.000000', '1.000000']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': '822f20d881eebeb9',\n",
       "  'bbox': ['0.000000', '1.000000', '0.181437', '0.963113']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': '15348e4f2c7ebe0f',\n",
       "  'bbox': ['0.087330', '0.999973', '0.600616', '0.999962']},\n",
       " {'dataset': 'validation',\n",
       "  'label': 'Bathtub',\n",
       "  'id': 'ddd3f96ff7f0bc78',\n",
       "  'bbox': ['0.251883', '0.976808', '0.460918', '1.000000']}]"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Now let's get bounding box information\n",
    "label_details = []\n",
    "for label_path in label_files:\n",
    "    dataset, label, img_id = os.path.splitext(label_path)[0].split(\"/\")\n",
    "    with open(label_path, \"r\") as file:\n",
    "        data = file.read().replace(\"\\n\", \"\").split(\",\")\n",
    "        XMin, XMax, YMin, YMax = data[1], data[2], data[3], data[4]\n",
    "        # bbox dimensions = XMin, XMax, YMin, YMax\n",
    "        #bbox = [float(i) for i in [data[1], data[2], data[3], data[4]]]\n",
    "        \n",
    "    label_detail = {\n",
    "        \"dataset\": dataset,\n",
    "        \"label\": label,\n",
    "        \"id\": img_id,\n",
    "        \"bbox\": [XMin, XMax, YMin, YMax]\n",
    "    }\n",
    "    label_details.append(label_detail)\n",
    "label_details"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Toilet,0.483064,0.583230,0.148464,0.254009\\n',\n",
       " 'Toilet,0.756245,0.872021,0.377751,0.570644\\n']"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "with open(\"validation/Toilet/c7312a8b436f5e90.txt\", \"r\") as file:\n",
    "    data = file.readlines()\n",
    "    \n",
    "data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "scrolled": true
   },
   "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>dataset</th>\n",
       "      <th>label</th>\n",
       "      <th>id</th>\n",
       "      <th>file_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>d60e1c25e87d8b45</td>\n",
       "      <td>validation/Toilet/d60e1c25e87d8b45.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>17f698ec871569ca</td>\n",
       "      <td>validation/Toilet/17f698ec871569ca.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>7c2825a3d8e0ed29</td>\n",
       "      <td>validation/Toilet/7c2825a3d8e0ed29.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>40a69f79da2aeaaa</td>\n",
       "      <td>validation/Toilet/40a69f79da2aeaaa.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>b2dc8e2437e8803f</td>\n",
       "      <td>validation/Toilet/b2dc8e2437e8803f.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>e6e05b56799fefba</td>\n",
       "      <td>validation/Toilet/e6e05b56799fefba.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>c1eefc708000b69f</td>\n",
       "      <td>validation/Toilet/c1eefc708000b69f.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>b3a65783709539de</td>\n",
       "      <td>validation/Toilet/b3a65783709539de.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>0d0719cfd8e417b7</td>\n",
       "      <td>validation/Toilet/0d0719cfd8e417b7.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>3454c87a14067798</td>\n",
       "      <td>validation/Toilet/3454c87a14067798.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>e108d68162990dfb</td>\n",
       "      <td>validation/Toilet/e108d68162990dfb.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>2f0a18a409d5a769</td>\n",
       "      <td>validation/Toilet/2f0a18a409d5a769.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>217b95a71cb220f8</td>\n",
       "      <td>validation/Toilet/217b95a71cb220f8.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>3a511ef0f68cc438</td>\n",
       "      <td>validation/Toilet/3a511ef0f68cc438.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>be84a3da54658167</td>\n",
       "      <td>validation/Toilet/be84a3da54658167.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>36d8c654fba5f337</td>\n",
       "      <td>validation/Toilet/36d8c654fba5f337.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>47ab0e73c33bdefd</td>\n",
       "      <td>validation/Toilet/47ab0e73c33bdefd.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>c7312a8b436f5e90</td>\n",
       "      <td>validation/Toilet/c7312a8b436f5e90.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>b821cc12eede040e</td>\n",
       "      <td>validation/Toilet/b821cc12eede040e.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>85ccc93d1f941931</td>\n",
       "      <td>validation/Toilet/85ccc93d1f941931.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>eaa75967cbc70ac1</td>\n",
       "      <td>validation/Toilet/eaa75967cbc70ac1.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>2ff2bb609a057f7a</td>\n",
       "      <td>validation/Toilet/2ff2bb609a057f7a.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>539e0871494dea5d</td>\n",
       "      <td>validation/Toilet/539e0871494dea5d.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>42298c9659ce6603</td>\n",
       "      <td>validation/Toilet/42298c9659ce6603.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>validation</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>f90eac5444f62b19</td>\n",
       "      <td>validation/Toilet/f90eac5444f62b19.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>validation</td>\n",
       "      <td>Bathtub</td>\n",
       "      <td>2f2039140c8b1f2b</td>\n",
       "      <td>validation/Bathtub/2f2039140c8b1f2b.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>validation</td>\n",
       "      <td>Bathtub</td>\n",
       "      <td>784bba5f12ee7c35</td>\n",
       "      <td>validation/Bathtub/784bba5f12ee7c35.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>validation</td>\n",
       "      <td>Bathtub</td>\n",
       "      <td>837b3d11ff02f116</td>\n",
       "      <td>validation/Bathtub/837b3d11ff02f116.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>validation</td>\n",
       "      <td>Bathtub</td>\n",
       "      <td>2f0a18a409d5a769</td>\n",
       "      <td>validation/Bathtub/2f0a18a409d5a769.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>validation</td>\n",
       "      <td>Bathtub</td>\n",
       "      <td>789cc0283c0e18af</td>\n",
       "      <td>validation/Bathtub/789cc0283c0e18af.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>validation</td>\n",
       "      <td>Bathtub</td>\n",
       "      <td>eccb080e57b2aac5</td>\n",
       "      <td>validation/Bathtub/eccb080e57b2aac5.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>validation</td>\n",
       "      <td>Bathtub</td>\n",
       "      <td>11dcac4ca5923a58</td>\n",
       "      <td>validation/Bathtub/11dcac4ca5923a58.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>validation</td>\n",
       "      <td>Bathtub</td>\n",
       "      <td>15348e4f2c7ebe0f</td>\n",
       "      <td>validation/Bathtub/15348e4f2c7ebe0f.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>validation</td>\n",
       "      <td>Bathtub</td>\n",
       "      <td>822f20d881eebeb9</td>\n",
       "      <td>validation/Bathtub/822f20d881eebeb9.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>validation</td>\n",
       "      <td>Bathtub</td>\n",
       "      <td>ddd3f96ff7f0bc78</td>\n",
       "      <td>validation/Bathtub/ddd3f96ff7f0bc78.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>validation</td>\n",
       "      <td>Bathtub</td>\n",
       "      <td>d7d469cb4c7e8cd2</td>\n",
       "      <td>validation/Bathtub/d7d469cb4c7e8cd2.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>validation</td>\n",
       "      <td>Bathtub</td>\n",
       "      <td>8e562fee6ff208f7</td>\n",
       "      <td>validation/Bathtub/8e562fee6ff208f7.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>validation</td>\n",
       "      <td>Bathtub</td>\n",
       "      <td>cd727c5e0d2cfc1b</td>\n",
       "      <td>validation/Bathtub/cd727c5e0d2cfc1b.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>validation</td>\n",
       "      <td>Bathtub</td>\n",
       "      <td>3f66fdf9688514d4</td>\n",
       "      <td>validation/Bathtub/3f66fdf9688514d4.jpg</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       dataset    label                id  \\\n",
       "0   validation   Toilet  d60e1c25e87d8b45   \n",
       "1   validation   Toilet  17f698ec871569ca   \n",
       "2   validation   Toilet  7c2825a3d8e0ed29   \n",
       "3   validation   Toilet  40a69f79da2aeaaa   \n",
       "4   validation   Toilet  b2dc8e2437e8803f   \n",
       "5   validation   Toilet  e6e05b56799fefba   \n",
       "6   validation   Toilet  c1eefc708000b69f   \n",
       "7   validation   Toilet  b3a65783709539de   \n",
       "8   validation   Toilet  0d0719cfd8e417b7   \n",
       "9   validation   Toilet  3454c87a14067798   \n",
       "10  validation   Toilet  e108d68162990dfb   \n",
       "11  validation   Toilet  2f0a18a409d5a769   \n",
       "12  validation   Toilet  217b95a71cb220f8   \n",
       "13  validation   Toilet  3a511ef0f68cc438   \n",
       "14  validation   Toilet  be84a3da54658167   \n",
       "15  validation   Toilet  36d8c654fba5f337   \n",
       "16  validation   Toilet  47ab0e73c33bdefd   \n",
       "17  validation   Toilet  c7312a8b436f5e90   \n",
       "18  validation   Toilet  b821cc12eede040e   \n",
       "19  validation   Toilet  85ccc93d1f941931   \n",
       "20  validation   Toilet  eaa75967cbc70ac1   \n",
       "21  validation   Toilet  2ff2bb609a057f7a   \n",
       "22  validation   Toilet  539e0871494dea5d   \n",
       "23  validation   Toilet  42298c9659ce6603   \n",
       "24  validation   Toilet  f90eac5444f62b19   \n",
       "25  validation  Bathtub  2f2039140c8b1f2b   \n",
       "26  validation  Bathtub  784bba5f12ee7c35   \n",
       "27  validation  Bathtub  837b3d11ff02f116   \n",
       "28  validation  Bathtub  2f0a18a409d5a769   \n",
       "29  validation  Bathtub  789cc0283c0e18af   \n",
       "30  validation  Bathtub  eccb080e57b2aac5   \n",
       "31  validation  Bathtub  11dcac4ca5923a58   \n",
       "32  validation  Bathtub  15348e4f2c7ebe0f   \n",
       "33  validation  Bathtub  822f20d881eebeb9   \n",
       "34  validation  Bathtub  ddd3f96ff7f0bc78   \n",
       "35  validation  Bathtub  d7d469cb4c7e8cd2   \n",
       "36  validation  Bathtub  8e562fee6ff208f7   \n",
       "37  validation  Bathtub  cd727c5e0d2cfc1b   \n",
       "38  validation  Bathtub  3f66fdf9688514d4   \n",
       "\n",
       "                                  file_name  \n",
       "0    validation/Toilet/d60e1c25e87d8b45.jpg  \n",
       "1    validation/Toilet/17f698ec871569ca.jpg  \n",
       "2    validation/Toilet/7c2825a3d8e0ed29.jpg  \n",
       "3    validation/Toilet/40a69f79da2aeaaa.jpg  \n",
       "4    validation/Toilet/b2dc8e2437e8803f.jpg  \n",
       "5    validation/Toilet/e6e05b56799fefba.jpg  \n",
       "6    validation/Toilet/c1eefc708000b69f.jpg  \n",
       "7    validation/Toilet/b3a65783709539de.jpg  \n",
       "8    validation/Toilet/0d0719cfd8e417b7.jpg  \n",
       "9    validation/Toilet/3454c87a14067798.jpg  \n",
       "10   validation/Toilet/e108d68162990dfb.jpg  \n",
       "11   validation/Toilet/2f0a18a409d5a769.jpg  \n",
       "12   validation/Toilet/217b95a71cb220f8.jpg  \n",
       "13   validation/Toilet/3a511ef0f68cc438.jpg  \n",
       "14   validation/Toilet/be84a3da54658167.jpg  \n",
       "15   validation/Toilet/36d8c654fba5f337.jpg  \n",
       "16   validation/Toilet/47ab0e73c33bdefd.jpg  \n",
       "17   validation/Toilet/c7312a8b436f5e90.jpg  \n",
       "18   validation/Toilet/b821cc12eede040e.jpg  \n",
       "19   validation/Toilet/85ccc93d1f941931.jpg  \n",
       "20   validation/Toilet/eaa75967cbc70ac1.jpg  \n",
       "21   validation/Toilet/2ff2bb609a057f7a.jpg  \n",
       "22   validation/Toilet/539e0871494dea5d.jpg  \n",
       "23   validation/Toilet/42298c9659ce6603.jpg  \n",
       "24   validation/Toilet/f90eac5444f62b19.jpg  \n",
       "25  validation/Bathtub/2f2039140c8b1f2b.jpg  \n",
       "26  validation/Bathtub/784bba5f12ee7c35.jpg  \n",
       "27  validation/Bathtub/837b3d11ff02f116.jpg  \n",
       "28  validation/Bathtub/2f0a18a409d5a769.jpg  \n",
       "29  validation/Bathtub/789cc0283c0e18af.jpg  \n",
       "30  validation/Bathtub/eccb080e57b2aac5.jpg  \n",
       "31  validation/Bathtub/11dcac4ca5923a58.jpg  \n",
       "32  validation/Bathtub/15348e4f2c7ebe0f.jpg  \n",
       "33  validation/Bathtub/822f20d881eebeb9.jpg  \n",
       "34  validation/Bathtub/ddd3f96ff7f0bc78.jpg  \n",
       "35  validation/Bathtub/d7d469cb4c7e8cd2.jpg  \n",
       "36  validation/Bathtub/8e562fee6ff208f7.jpg  \n",
       "37  validation/Bathtub/cd727c5e0d2cfc1b.jpg  \n",
       "38  validation/Bathtub/3f66fdf9688514d4.jpg  "
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame(img_details)\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# FIGURED IT OUT\n",
    "\n",
    "Why am I doing such hard work manipulating images when I could just use their own annotation files?\n",
    "\n",
    "Why give a sh*t about the text files for labels when I've already got validation-annotations-bbox.csv... and the same with train-annotations-bbox.csv... derp. Time to use both of these to my advantage and just manipulate the images to be in the folders I want them in.\n",
    "\n",
    "I should write a function which:\n",
    "- Gets my image IDs\n",
    "- Matches them with their details in the annotations csv \n",
    "- Saves their file name"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "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>ImageID</th>\n",
       "      <th>Source</th>\n",
       "      <th>LabelName</th>\n",
       "      <th>Confidence</th>\n",
       "      <th>XMin</th>\n",
       "      <th>XMax</th>\n",
       "      <th>YMin</th>\n",
       "      <th>YMax</th>\n",
       "      <th>IsOccluded</th>\n",
       "      <th>IsTruncated</th>\n",
       "      <th>IsGroupOf</th>\n",
       "      <th>IsDepiction</th>\n",
       "      <th>IsInside</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0001eeaf4aed83f9</td>\n",
       "      <td>freeform</td>\n",
       "      <td>/m/0cmf2</td>\n",
       "      <td>1</td>\n",
       "      <td>0.022464</td>\n",
       "      <td>0.964178</td>\n",
       "      <td>0.070656</td>\n",
       "      <td>0.800164</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>000595fe6fee6369</td>\n",
       "      <td>freeform</td>\n",
       "      <td>/m/02wbm</td>\n",
       "      <td>1</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000233</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>000595fe6fee6369</td>\n",
       "      <td>freeform</td>\n",
       "      <td>/m/02xwb</td>\n",
       "      <td>1</td>\n",
       "      <td>0.141030</td>\n",
       "      <td>0.180277</td>\n",
       "      <td>0.676262</td>\n",
       "      <td>0.732455</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>000595fe6fee6369</td>\n",
       "      <td>freeform</td>\n",
       "      <td>/m/02xwb</td>\n",
       "      <td>1</td>\n",
       "      <td>0.213781</td>\n",
       "      <td>0.253028</td>\n",
       "      <td>0.298764</td>\n",
       "      <td>0.354956</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>000595fe6fee6369</td>\n",
       "      <td>freeform</td>\n",
       "      <td>/m/02xwb</td>\n",
       "      <td>1</td>\n",
       "      <td>0.232926</td>\n",
       "      <td>0.288447</td>\n",
       "      <td>0.488954</td>\n",
       "      <td>0.545146</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>204616</th>\n",
       "      <td>ffff21932da3ed01</td>\n",
       "      <td>freeform</td>\n",
       "      <td>/m/03fp41</td>\n",
       "      <td>1</td>\n",
       "      <td>0.177790</td>\n",
       "      <td>0.710296</td>\n",
       "      <td>0.412302</td>\n",
       "      <td>0.578570</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>204617</th>\n",
       "      <td>ffff21932da3ed01</td>\n",
       "      <td>freeform</td>\n",
       "      <td>/m/05s2s</td>\n",
       "      <td>1</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.031963</td>\n",
       "      <td>0.502994</td>\n",
       "      <td>0.562275</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>204618</th>\n",
       "      <td>ffff21932da3ed01</td>\n",
       "      <td>freeform</td>\n",
       "      <td>/m/0c9ph5</td>\n",
       "      <td>1</td>\n",
       "      <td>0.323775</td>\n",
       "      <td>0.409382</td>\n",
       "      <td>0.464495</td>\n",
       "      <td>0.554111</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>204619</th>\n",
       "      <td>ffff21932da3ed01</td>\n",
       "      <td>freeform</td>\n",
       "      <td>/m/0c9ph5</td>\n",
       "      <td>1</td>\n",
       "      <td>0.540223</td>\n",
       "      <td>0.624863</td>\n",
       "      <td>0.493633</td>\n",
       "      <td>0.577892</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>204620</th>\n",
       "      <td>ffff21932da3ed01</td>\n",
       "      <td>freeform</td>\n",
       "      <td>/m/0cgh4</td>\n",
       "      <td>1</td>\n",
       "      <td>0.002521</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.998685</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>204621 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                 ImageID    Source  LabelName  Confidence      XMin      XMax  \\\n",
       "0       0001eeaf4aed83f9  freeform   /m/0cmf2           1  0.022464  0.964178   \n",
       "1       000595fe6fee6369  freeform   /m/02wbm           1  0.000000  1.000000   \n",
       "2       000595fe6fee6369  freeform   /m/02xwb           1  0.141030  0.180277   \n",
       "3       000595fe6fee6369  freeform   /m/02xwb           1  0.213781  0.253028   \n",
       "4       000595fe6fee6369  freeform   /m/02xwb           1  0.232926  0.288447   \n",
       "...                  ...       ...        ...         ...       ...       ...   \n",
       "204616  ffff21932da3ed01  freeform  /m/03fp41           1  0.177790  0.710296   \n",
       "204617  ffff21932da3ed01  freeform   /m/05s2s           1  0.000000  0.031963   \n",
       "204618  ffff21932da3ed01  freeform  /m/0c9ph5           1  0.323775  0.409382   \n",
       "204619  ffff21932da3ed01  freeform  /m/0c9ph5           1  0.540223  0.624863   \n",
       "204620  ffff21932da3ed01  freeform   /m/0cgh4           1  0.002521  1.000000   \n",
       "\n",
       "            YMin      YMax  IsOccluded  IsTruncated  IsGroupOf  IsDepiction  \\\n",
       "0       0.070656  0.800164           0            0          0            0   \n",
       "1       0.000233  1.000000           0            0          1            0   \n",
       "2       0.676262  0.732455           0            0          0            0   \n",
       "3       0.298764  0.354956           1            0          0            0   \n",
       "4       0.488954  0.545146           1            0          0            0   \n",
       "...          ...       ...         ...          ...        ...          ...   \n",
       "204616  0.412302  0.578570           0            0          1            0   \n",
       "204617  0.502994  0.562275           1            1          0            0   \n",
       "204618  0.464495  0.554111           0            0          1            0   \n",
       "204619  0.493633  0.577892           1            0          1            0   \n",
       "204620  0.000000  0.998685           0            0          0            0   \n",
       "\n",
       "        IsInside  \n",
       "0              0  \n",
       "1              0  \n",
       "2              0  \n",
       "3              0  \n",
       "4              0  \n",
       "...          ...  \n",
       "204616         0  \n",
       "204617         0  \n",
       "204618         0  \n",
       "204619         0  \n",
       "204620         1  \n",
       "\n",
       "[204621 rows x 13 columns]"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "val_annot = pd.read_csv(\"validation-annotations-bbox.csv\")\n",
    "val_annot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "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>id</th>\n",
       "      <th>ClassName</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>/m/011k07</td>\n",
       "      <td>Tortoise</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>/m/011q46kg</td>\n",
       "      <td>Container</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>/m/012074</td>\n",
       "      <td>Magpie</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>/m/0120dh</td>\n",
       "      <td>Sea turtle</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>/m/01226z</td>\n",
       "      <td>Football</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>596</th>\n",
       "      <td>/m/0qmmr</td>\n",
       "      <td>Wheelchair</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>597</th>\n",
       "      <td>/m/0wdt60w</td>\n",
       "      <td>Rugby ball</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>598</th>\n",
       "      <td>/m/0xfy</td>\n",
       "      <td>Armadillo</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>599</th>\n",
       "      <td>/m/0xzly</td>\n",
       "      <td>Maracas</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>600</th>\n",
       "      <td>/m/0zvk5</td>\n",
       "      <td>Helmet</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>601 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              id   ClassName\n",
       "0      /m/011k07    Tortoise\n",
       "1    /m/011q46kg   Container\n",
       "2      /m/012074      Magpie\n",
       "3      /m/0120dh  Sea turtle\n",
       "4      /m/01226z    Football\n",
       "..           ...         ...\n",
       "596     /m/0qmmr  Wheelchair\n",
       "597   /m/0wdt60w  Rugby ball\n",
       "598      /m/0xfy   Armadillo\n",
       "599     /m/0xzly     Maracas\n",
       "600     /m/0zvk5      Helmet\n",
       "\n",
       "[601 rows x 2 columns]"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "classes = pd.read_csv(\"class-descriptions-boxable.csv\", names=[\"id\", \"ClassName\"])\n",
    "classes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 29.7 ms, sys: 3.32 ms, total: 33 ms\n",
      "Wall time: 31.8 ms\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0           Airplane\n",
       "1               Food\n",
       "2              Fruit\n",
       "3              Fruit\n",
       "4              Fruit\n",
       "             ...    \n",
       "204616    Houseplant\n",
       "204617         Plant\n",
       "204618        Flower\n",
       "204619        Flower\n",
       "204620      Building\n",
       "Name: ClassName, Length: 204621, dtype: object"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "# Create classname column on annotations which converts label codes to string labels\n",
    "val_annot[\"ClassName\"] = val_annot[\"LabelName\"].map(classes.set_index(\"id\")[\"ClassName\"])\n",
    "val_annot[\"ClassName\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Move all images into one file (images)\n",
    "Since I've already got a key (the annotations CSV), I'm going to move all the images into one file and manipulate them later."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['validation/Toilet/d60e1c25e87d8b45.jpg',\n",
       " 'validation/Toilet/17f698ec871569ca.jpg',\n",
       " 'validation/Toilet/7c2825a3d8e0ed29.jpg']"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Take these and move them into one file\n",
    "img_files[:3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Make directory with images\n",
    "os.mkdir(\"images\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Copy images files to images\n",
    "import shutil\n",
    "for image in img_files:\n",
    "    shutil.copy2(image, \"images\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(38, 39)"
      ]
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(os.listdir(\"images\")), len(img_files)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "38"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "len(np.unique([i.split(\"/\")[2] for i in img_files]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['d60e1c25e87d8b45.jpg',\n",
       " '2f2039140c8b1f2b.jpg',\n",
       " '17f698ec871569ca.jpg',\n",
       " '7c2825a3d8e0ed29.jpg',\n",
       " '40a69f79da2aeaaa.jpg']"
      ]
     },
     "execution_count": 115,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "os.listdir(\"images\")[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['d60e1c25e87d8b45',\n",
       " '2f2039140c8b1f2b',\n",
       " '17f698ec871569ca',\n",
       " '7c2825a3d8e0ed29',\n",
       " '40a69f79da2aeaaa',\n",
       " 'b2dc8e2437e8803f',\n",
       " 'e6e05b56799fefba',\n",
       " 'c1eefc708000b69f',\n",
       " 'b3a65783709539de',\n",
       " '0d0719cfd8e417b7',\n",
       " '3454c87a14067798',\n",
       " '784bba5f12ee7c35',\n",
       " 'e108d68162990dfb',\n",
       " '837b3d11ff02f116',\n",
       " '2f0a18a409d5a769',\n",
       " '217b95a71cb220f8',\n",
       " '3a511ef0f68cc438',\n",
       " 'be84a3da54658167',\n",
       " '36d8c654fba5f337',\n",
       " '47ab0e73c33bdefd',\n",
       " '789cc0283c0e18af',\n",
       " 'c7312a8b436f5e90',\n",
       " 'eccb080e57b2aac5',\n",
       " '11dcac4ca5923a58',\n",
       " 'b821cc12eede040e',\n",
       " '85ccc93d1f941931',\n",
       " '15348e4f2c7ebe0f',\n",
       " '822f20d881eebeb9',\n",
       " 'eaa75967cbc70ac1',\n",
       " '2ff2bb609a057f7a',\n",
       " 'ddd3f96ff7f0bc78',\n",
       " 'd7d469cb4c7e8cd2',\n",
       " '539e0871494dea5d',\n",
       " '42298c9659ce6603',\n",
       " '8e562fee6ff208f7',\n",
       " 'cd727c5e0d2cfc1b',\n",
       " 'f90eac5444f62b19',\n",
       " '3f66fdf9688514d4']"
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Make list of images we have and split them off the dataframe with all the label information\n",
    "my_images = [os.path.splitext(img_name)[0] for img_name in os.listdir(\"images\")]\n",
    "my_images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Toilet',\n",
       " 'Swimming_pool',\n",
       " 'Bed',\n",
       " 'Biliard_table',\n",
       " 'Sink',\n",
       " 'Fountain',\n",
       " 'Oven',\n",
       " 'Ceiling_fan',\n",
       " 'Television',\n",
       " 'Microwave_oven',\n",
       " 'Gas_stove',\n",
       " 'Refrigerator',\n",
       " 'Kitchen_&_dining_room_table',\n",
       " 'Washing_machine',\n",
       " 'Bathtub',\n",
       " 'Stairs',\n",
       " 'Fireplace',\n",
       " 'Pillow',\n",
       " 'Mirror',\n",
       " 'Shower',\n",
       " 'Couch',\n",
       " 'Countertop',\n",
       " 'Coffeemaker',\n",
       " 'Dishwasher',\n",
       " 'Sofa_bed',\n",
       " 'Tree_house',\n",
       " 'Towel',\n",
       " 'Porch',\n",
       " 'Wine_rack',\n",
       " 'Jacuzzi']"
      ]
     },
     "execution_count": 122,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "subset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "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>ImageID</th>\n",
       "      <th>Source</th>\n",
       "      <th>LabelName</th>\n",
       "      <th>Confidence</th>\n",
       "      <th>XMin</th>\n",
       "      <th>XMax</th>\n",
       "      <th>YMin</th>\n",
       "      <th>YMax</th>\n",
       "      <th>IsOccluded</th>\n",
       "      <th>IsTruncated</th>\n",
       "      <th>IsGroupOf</th>\n",
       "      <th>IsDepiction</th>\n",
       "      <th>IsInside</th>\n",
       "      <th>ClassName</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>10797</th>\n",
       "      <td>0d0719cfd8e417b7</td>\n",
       "      <td>freeform</td>\n",
       "      <td>/m/09g1w</td>\n",
       "      <td>1</td>\n",
       "      <td>0.093371</td>\n",
       "      <td>0.986232</td>\n",
       "      <td>0.189984</td>\n",
       "      <td>0.965806</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Toilet</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14593</th>\n",
       "      <td>11dcac4ca5923a58</td>\n",
       "      <td>freeform</td>\n",
       "      <td>/m/03dnzn</td>\n",
       "      <td>1</td>\n",
       "      <td>0.360645</td>\n",
       "      <td>0.999988</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Bathtub</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17325</th>\n",
       "      <td>15348e4f2c7ebe0f</td>\n",
       "      <td>freeform</td>\n",
       "      <td>/m/03dnzn</td>\n",
       "      <td>1</td>\n",
       "      <td>0.087330</td>\n",
       "      <td>0.999973</td>\n",
       "      <td>0.600616</td>\n",
       "      <td>0.999962</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Bathtub</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17329</th>\n",
       "      <td>15348e4f2c7ebe0f</td>\n",
       "      <td>freeform</td>\n",
       "      <td>/m/065h6l</td>\n",
       "      <td>1</td>\n",
       "      <td>0.090833</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.600928</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Jacuzzi</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19543</th>\n",
       "      <td>17f698ec871569ca</td>\n",
       "      <td>freeform</td>\n",
       "      <td>/m/09g1w</td>\n",
       "      <td>1</td>\n",
       "      <td>0.160562</td>\n",
       "      <td>0.475538</td>\n",
       "      <td>0.181890</td>\n",
       "      <td>0.847251</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Toilet</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                ImageID    Source  LabelName  Confidence      XMin      XMax  \\\n",
       "10797  0d0719cfd8e417b7  freeform   /m/09g1w           1  0.093371  0.986232   \n",
       "14593  11dcac4ca5923a58  freeform  /m/03dnzn           1  0.360645  0.999988   \n",
       "17325  15348e4f2c7ebe0f  freeform  /m/03dnzn           1  0.087330  0.999973   \n",
       "17329  15348e4f2c7ebe0f  freeform  /m/065h6l           1  0.090833  1.000000   \n",
       "19543  17f698ec871569ca  freeform   /m/09g1w           1  0.160562  0.475538   \n",
       "\n",
       "           YMin      YMax  IsOccluded  IsTruncated  IsGroupOf  IsDepiction  \\\n",
       "10797  0.189984  0.965806           0            0          0            0   \n",
       "14593  0.000000  1.000000           0            1          0            0   \n",
       "17325  0.600616  0.999962           0            0          0            0   \n",
       "17329  0.600928  1.000000           0            0          0            0   \n",
       "19543  0.181890  0.847251           0            0          0            0   \n",
       "\n",
       "       IsInside ClassName  \n",
       "10797         0    Toilet  \n",
       "14593         0   Bathtub  \n",
       "17325         0   Bathtub  \n",
       "17329         0   Jacuzzi  \n",
       "19543         0    Toilet  "
      ]
     },
     "execution_count": 125,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Make sure we only get the images we're concerned about\n",
    "my_images_df = val_annot[val_annot[\"ImageID\"].isin(my_images) & val_annot[\"ClassName\"].isin(subset)]\n",
    "my_images_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([3, 0, 0, 1, 3, 3, 3, 0, 3, 0, 3, 3, 3, 3, 2, 3, 0, 3, 3, 3, 3, 0,\n",
       "       0, 3, 0, 1, 0, 3, 0, 0, 3, 3, 3, 3, 3, 2, 2, 2, 3, 3, 0, 2, 3, 0,\n",
       "       1, 0, 3, 3, 3, 0, 1, 3], dtype=int8)"
      ]
     },
     "execution_count": 205,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.Categorical(my_images_df.ClassName).codes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['Sink', 'Sink', 'Sink', 'Toilet', 'Toilet'], dtype=object)"
      ]
     },
     "execution_count": 143,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "my_images_df[my_images_df[\"ImageID\"] == \"c7312a8b436f5e90\"][\"ClassName\"].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.238995, 0.372361, 0.778648, 0.871384],\n",
       "       [0.461353, 0.615221, 0.322836, 0.441359],\n",
       "       [0.782433, 0.921636, 0.850595, 0.936918],\n",
       "       [0.483064, 0.58323 , 0.148464, 0.254009],\n",
       "       [0.756245, 0.872021, 0.377751, 0.570644]])"
      ]
     },
     "execution_count": 148,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "my_images_df[my_images_df[\"ImageID\"] == \"c7312a8b436f5e90\"][[\"XMin\", \"XMax\", \"YMin\", \"YMax\"]].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "52"
      ]
     },
     "execution_count": 137,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(my_images_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "c7312a8b436f5e90    5\n",
       "15348e4f2c7ebe0f    2\n",
       "3a511ef0f68cc438    2\n",
       "d7d469cb4c7e8cd2    2\n",
       "d60e1c25e87d8b45    2\n",
       "3454c87a14067798    2\n",
       "17f698ec871569ca    2\n",
       "eccb080e57b2aac5    2\n",
       "822f20d881eebeb9    2\n",
       "2f0a18a409d5a769    2\n",
       "8e562fee6ff208f7    2\n",
       "e108d68162990dfb    1\n",
       "539e0871494dea5d    1\n",
       "be84a3da54658167    1\n",
       "b821cc12eede040e    1\n",
       "ddd3f96ff7f0bc78    1\n",
       "b3a65783709539de    1\n",
       "eaa75967cbc70ac1    1\n",
       "85ccc93d1f941931    1\n",
       "7c2825a3d8e0ed29    1\n",
       "f90eac5444f62b19    1\n",
       "2ff2bb609a057f7a    1\n",
       "789cc0283c0e18af    1\n",
       "784bba5f12ee7c35    1\n",
       "2f2039140c8b1f2b    1\n",
       "11dcac4ca5923a58    1\n",
       "40a69f79da2aeaaa    1\n",
       "e6e05b56799fefba    1\n",
       "c1eefc708000b69f    1\n",
       "36d8c654fba5f337    1\n",
       "42298c9659ce6603    1\n",
       "3f66fdf9688514d4    1\n",
       "0d0719cfd8e417b7    1\n",
       "b2dc8e2437e8803f    1\n",
       "cd727c5e0d2cfc1b    1\n",
       "47ab0e73c33bdefd    1\n",
       "217b95a71cb220f8    1\n",
       "837b3d11ff02f116    1\n",
       "Name: ImageID, dtype: int64"
      ]
     },
     "execution_count": 140,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "my_images_df[\"ImageID\"].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['d60e1c25e87d8b45',\n",
       " '2f2039140c8b1f2b',\n",
       " '17f698ec871569ca',\n",
       " '7c2825a3d8e0ed29',\n",
       " '40a69f79da2aeaaa',\n",
       " 'b2dc8e2437e8803f',\n",
       " 'e6e05b56799fefba',\n",
       " 'c1eefc708000b69f',\n",
       " 'b3a65783709539de',\n",
       " '0d0719cfd8e417b7',\n",
       " '3454c87a14067798',\n",
       " '784bba5f12ee7c35',\n",
       " 'e108d68162990dfb',\n",
       " '837b3d11ff02f116',\n",
       " '2f0a18a409d5a769',\n",
       " '217b95a71cb220f8',\n",
       " '3a511ef0f68cc438',\n",
       " 'be84a3da54658167',\n",
       " '36d8c654fba5f337',\n",
       " '47ab0e73c33bdefd',\n",
       " '789cc0283c0e18af',\n",
       " 'c7312a8b436f5e90',\n",
       " 'eccb080e57b2aac5',\n",
       " '11dcac4ca5923a58',\n",
       " 'b821cc12eede040e',\n",
       " '85ccc93d1f941931',\n",
       " '15348e4f2c7ebe0f',\n",
       " '822f20d881eebeb9',\n",
       " 'eaa75967cbc70ac1',\n",
       " '2ff2bb609a057f7a',\n",
       " 'ddd3f96ff7f0bc78',\n",
       " 'd7d469cb4c7e8cd2',\n",
       " '539e0871494dea5d',\n",
       " '42298c9659ce6603',\n",
       " '8e562fee6ff208f7',\n",
       " 'cd727c5e0d2cfc1b',\n",
       " 'f90eac5444f62b19',\n",
       " '3f66fdf9688514d4']"
      ]
     },
     "execution_count": 153,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "my_images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "metadata": {},
   "outputs": [
    {
     "ename": "IndexError",
     "evalue": "list assignment index out of range",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mIndexError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-154-f68825abac0c>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0mlisty_list\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m \u001b[0mlisty_list\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;31mIndexError\u001b[0m: list assignment index out of range"
     ]
    }
   ],
   "source": [
    "listy_list = []"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "# Create images labels setup in detectron2 style\n",
    "img_dicts = []\n",
    "for i, img in enumerate(my_images):\n",
    "    file_path = \"images/\" + img + \".jpg\"\n",
    "    img_data = my_images_df[my_images_df[\"ImageID\"] == img]\n",
    "    img_label = img_data[\"ClassName\"].values\n",
    "    bboxes = img_data[[\"XMin\", \"XMax\", \"YMin\", \"YMax\"]].values\n",
    "    img_dict = {\"file_path\": file_path,\n",
    "                \"img_label\": img_label,\n",
    "                \"bboxes\": bboxes}\n",
    "    img_dicts.append(img_dict)\n",
    "img_dicts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "index                    172106\n",
      "ImageID        d60e1c25e87d8b45\n",
      "Source                 freeform\n",
      "LabelName             /m/0130jx\n",
      "Confidence                    1\n",
      "XMin                          0\n",
      "XMax                   0.823265\n",
      "YMin                   0.315739\n",
      "YMax                   0.975746\n",
      "IsOccluded                    0\n",
      "IsTruncated                   0\n",
      "IsGroupOf                     0\n",
      "IsDepiction                   0\n",
      "IsInside                      0\n",
      "ClassName                  Sink\n",
      "Name: 0, dtype: object\n",
      "index                    172112\n",
      "ImageID        d60e1c25e87d8b45\n",
      "Source                 freeform\n",
      "LabelName              /m/09g1w\n",
      "Confidence                    1\n",
      "XMin                   0.405653\n",
      "XMax                          1\n",
      "YMin                   0.146613\n",
      "YMax                   0.626838\n",
      "IsOccluded                    1\n",
      "IsTruncated                   0\n",
      "IsGroupOf                     0\n",
      "IsDepiction                   0\n",
      "IsInside                      0\n",
      "ClassName                Toilet\n",
      "Name: 1, dtype: object\n",
      "CPU times: user 4.91 ms, sys: 280 µs, total: 5.19 ms\n",
      "Wall time: 5.21 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "multiple = my_images_df[my_images_df[\"ImageID\"] == \"d60e1c25e87d8b45\"].reset_index()\n",
    "for i in range(len(multiple)):\n",
    "    print(multiple.loc[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 220,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1.94 ms, sys: 200 µs, total: 2.14 ms\n",
      "Wall time: 1.98 ms\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/daniel/Desktop/airbnb-project/env/lib/python3.7/site-packages/pandas/core/indexing.py:965: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  self.obj[item] = s\n"
     ]
    },
    {
     "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>ImageID</th>\n",
       "      <th>Source</th>\n",
       "      <th>LabelName</th>\n",
       "      <th>Confidence</th>\n",
       "      <th>XMin</th>\n",
       "      <th>XMax</th>\n",
       "      <th>YMin</th>\n",
       "      <th>YMax</th>\n",
       "      <th>IsOccluded</th>\n",
       "      <th>IsTruncated</th>\n",
       "      <th>IsGroupOf</th>\n",
       "      <th>IsDepiction</th>\n",
       "      <th>IsInside</th>\n",
       "      <th>ClassName</th>\n",
       "      <th>ClassID</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>10797</th>\n",
       "      <td>0d0719cfd8e417b7</td>\n",
       "      <td>freeform</td>\n",
       "      <td>/m/09g1w</td>\n",
       "      <td>1</td>\n",
       "      <td>0.093371</td>\n",
       "      <td>0.986232</td>\n",
       "      <td>0.189984</td>\n",
       "      <td>0.965806</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14593</th>\n",
       "      <td>11dcac4ca5923a58</td>\n",
       "      <td>freeform</td>\n",
       "      <td>/m/03dnzn</td>\n",
       "      <td>1</td>\n",
       "      <td>0.360645</td>\n",
       "      <td>0.999988</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Bathtub</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17325</th>\n",
       "      <td>15348e4f2c7ebe0f</td>\n",
       "      <td>freeform</td>\n",
       "      <td>/m/03dnzn</td>\n",
       "      <td>1</td>\n",
       "      <td>0.087330</td>\n",
       "      <td>0.999973</td>\n",
       "      <td>0.600616</td>\n",
       "      <td>0.999962</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Bathtub</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17329</th>\n",
       "      <td>15348e4f2c7ebe0f</td>\n",
       "      <td>freeform</td>\n",
       "      <td>/m/065h6l</td>\n",
       "      <td>1</td>\n",
       "      <td>0.090833</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.600928</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Jacuzzi</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19543</th>\n",
       "      <td>17f698ec871569ca</td>\n",
       "      <td>freeform</td>\n",
       "      <td>/m/09g1w</td>\n",
       "      <td>1</td>\n",
       "      <td>0.160562</td>\n",
       "      <td>0.475538</td>\n",
       "      <td>0.181890</td>\n",
       "      <td>0.847251</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Toilet</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                ImageID    Source  LabelName  Confidence      XMin      XMax  \\\n",
       "10797  0d0719cfd8e417b7  freeform   /m/09g1w           1  0.093371  0.986232   \n",
       "14593  11dcac4ca5923a58  freeform  /m/03dnzn           1  0.360645  0.999988   \n",
       "17325  15348e4f2c7ebe0f  freeform  /m/03dnzn           1  0.087330  0.999973   \n",
       "17329  15348e4f2c7ebe0f  freeform  /m/065h6l           1  0.090833  1.000000   \n",
       "19543  17f698ec871569ca  freeform   /m/09g1w           1  0.160562  0.475538   \n",
       "\n",
       "           YMin      YMax  IsOccluded  IsTruncated  IsGroupOf  IsDepiction  \\\n",
       "10797  0.189984  0.965806           0            0          0            0   \n",
       "14593  0.000000  1.000000           0            1          0            0   \n",
       "17325  0.600616  0.999962           0            0          0            0   \n",
       "17329  0.600928  1.000000           0            0          0            0   \n",
       "19543  0.181890  0.847251           0            0          0            0   \n",
       "\n",
       "       IsInside ClassName  ClassID  \n",
       "10797         0    Toilet        3  \n",
       "14593         0   Bathtub        0  \n",
       "17325         0   Bathtub        0  \n",
       "17329         0   Jacuzzi        1  \n",
       "19543         0    Toilet        3  "
      ]
     },
     "execution_count": 220,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "category_ids = pd.Categorical(my_images_df[\"ClassName\"]).codes\n",
    "my_images_df.loc[:, \"ClassID\"] = category_ids\n",
    "my_images_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 219,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([3, 0, 0, 1, 3, 3, 3, 0, 3, 0, 3, 3, 3, 3, 2, 3, 0, 3, 3, 3, 3, 0,\n",
       "       0, 3, 0, 1, 0, 3, 0, 0, 3, 3, 3, 3, 3, 2, 2, 2, 3, 3, 0, 2, 3, 0,\n",
       "       1, 0, 3, 3, 3, 0, 1, 3], dtype=int8)"
      ]
     },
     "execution_count": 219,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "category_ids"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "metadata": {},
   "outputs": [],
   "source": [
    "targ = my_images_df[my_images_df[\"ImageID\"] == \"d60e1c25e87d8b45\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 208,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 1], dtype=int8)"
      ]
     },
     "execution_count": 208,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.Categorical(targ.ClassName).codes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'d60e1c25e87d8b45'"
      ]
     },
     "execution_count": 170,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "multiple.loc[0][\"ImageID\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/jpeg": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 191,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import Image\n",
    "Image(\"images/d60e1c25e87d8b45.jpg\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 225,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'filename': 'images/d60e1c25e87d8b45.jpg',\n",
       " 'image_id': 'd60e1c25e87d8b45',\n",
       " 'height': 0,\n",
       " 'width': 0,\n",
       " 'annotations': [{'bbox': [array([0.0, 0.8232649999999999, 0.315739, 0.9757459999999999],\n",
       "          dtype=object)],\n",
       "   'bbox_mode': 0,\n",
       "   'category_id': 2},\n",
       "  {'bbox': [array([0.40565300000000004, 1.0, 0.14661300000000002, 0.626838],\n",
       "          dtype=object)],\n",
       "   'bbox_mode': 0,\n",
       "   'category_id': 3}]}"
      ]
     },
     "execution_count": 225,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "img_dicts[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 224,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 213 ms, sys: 6.17 ms, total: 219 ms\n",
      "Wall time: 243 ms\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[{'filename': 'images/d60e1c25e87d8b45.jpg',\n",
       "  'image_id': 'd60e1c25e87d8b45',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.0, 0.8232649999999999, 0.315739, 0.9757459999999999],\n",
       "           dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 2},\n",
       "   {'bbox': [array([0.40565300000000004, 1.0, 0.14661300000000002, 0.626838],\n",
       "           dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/2f2039140c8b1f2b.jpg',\n",
       "  'image_id': '2f2039140c8b1f2b',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.0, 1.0, 0.0, 1.0], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 0}]},\n",
       " {'filename': 'images/17f698ec871569ca.jpg',\n",
       "  'image_id': '17f698ec871569ca',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.16056199999999998, 0.475538, 0.18189, 0.847251], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3},\n",
       "   {'bbox': [array([0.6201770000000001, 0.930982, 0.18248399999999998, 0.813342],\n",
       "           dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/7c2825a3d8e0ed29.jpg',\n",
       "  'image_id': '7c2825a3d8e0ed29',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.20705700000000002, 0.41516099999999995, 0.809775, 0.999841],\n",
       "           dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/40a69f79da2aeaaa.jpg',\n",
       "  'image_id': '40a69f79da2aeaaa',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.014169999999999999, 0.473459, 0.348211, 0.999983], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/b2dc8e2437e8803f.jpg',\n",
       "  'image_id': 'b2dc8e2437e8803f',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.337954, 0.695044, 0.258797, 0.857483], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/e6e05b56799fefba.jpg',\n",
       "  'image_id': 'e6e05b56799fefba',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.059556, 0.956108, 0.0, 0.5071829999999999], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/c1eefc708000b69f.jpg',\n",
       "  'image_id': 'c1eefc708000b69f',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.201985, 0.772443, 0.097975, 0.983168], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/b3a65783709539de.jpg',\n",
       "  'image_id': 'b3a65783709539de',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.18490399999999999, 0.8468950000000001, 0.060172,\n",
       "            0.9612219999999999], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/0d0719cfd8e417b7.jpg',\n",
       "  'image_id': '0d0719cfd8e417b7',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.09337100000000001, 0.986232, 0.189984, 0.9658059999999999],\n",
       "           dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/3454c87a14067798.jpg',\n",
       "  'image_id': '3454c87a14067798',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.077027, 0.306008, 0.811938, 1.0], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3},\n",
       "   {'bbox': [array([0.583791, 0.820996, 0.810397, 0.999875], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/784bba5f12ee7c35.jpg',\n",
       "  'image_id': '784bba5f12ee7c35',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.0, 1.0, 0.0, 1.0], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 0}]},\n",
       " {'filename': 'images/e108d68162990dfb.jpg',\n",
       "  'image_id': 'e108d68162990dfb',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.36077, 0.656517, 0.41038199999999997, 0.8860370000000001],\n",
       "           dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/837b3d11ff02f116.jpg',\n",
       "  'image_id': '837b3d11ff02f116',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.0, 0.669035, 0.37294099999999997, 1.0], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 0}]},\n",
       " {'filename': 'images/2f0a18a409d5a769.jpg',\n",
       "  'image_id': '2f0a18a409d5a769',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.157351, 1.0, 0.328732, 0.927573], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 0},\n",
       "   {'bbox': [array([8e-06, 0.5462239999999999, 0.349488, 0.99998], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/217b95a71cb220f8.jpg',\n",
       "  'image_id': '217b95a71cb220f8',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.438669, 0.602654, 0.44908100000000006, 0.532643], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/3a511ef0f68cc438.jpg',\n",
       "  'image_id': '3a511ef0f68cc438',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.22748000000000002, 0.493562, 0.625067, 0.78379], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 2},\n",
       "   {'bbox': [array([0.124107, 0.338616, 0.755953, 0.9788020000000001], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/be84a3da54658167.jpg',\n",
       "  'image_id': 'be84a3da54658167',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.246185, 0.776473, 0.72837, 1.0], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/36d8c654fba5f337.jpg',\n",
       "  'image_id': '36d8c654fba5f337',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.063034, 0.805088, 0.0, 0.697051], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/47ab0e73c33bdefd.jpg',\n",
       "  'image_id': '47ab0e73c33bdefd',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.197381, 0.9225559999999999, 0.468605, 0.829202], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/789cc0283c0e18af.jpg',\n",
       "  'image_id': '789cc0283c0e18af',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.039686, 0.9487180000000001, 0.11958099999999999, 1.0],\n",
       "           dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 0}]},\n",
       " {'filename': 'images/c7312a8b436f5e90.jpg',\n",
       "  'image_id': 'c7312a8b436f5e90',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.23899499999999999, 0.372361, 0.778648, 0.871384], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 2},\n",
       "   {'bbox': [array([0.46135299999999996, 0.615221, 0.322836, 0.441359], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 2},\n",
       "   {'bbox': [array([0.782433, 0.921636, 0.8505950000000001, 0.936918], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 2},\n",
       "   {'bbox': [array([0.48306400000000005, 0.58323, 0.148464, 0.25400900000000004],\n",
       "           dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3},\n",
       "   {'bbox': [array([0.756245, 0.8720209999999999, 0.377751, 0.5706439999999999],\n",
       "           dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/eccb080e57b2aac5.jpg',\n",
       "  'image_id': 'eccb080e57b2aac5',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.30334099999999997, 0.9058290000000001, 0.47085200000000005,\n",
       "            0.999948], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 0},\n",
       "   {'bbox': [array([0.30196300000000004, 0.90845, 0.472029, 0.999245], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 1}]},\n",
       " {'filename': 'images/11dcac4ca5923a58.jpg',\n",
       "  'image_id': '11dcac4ca5923a58',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.360645, 0.9999879999999999, 0.0, 1.0], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 0}]},\n",
       " {'filename': 'images/b821cc12eede040e.jpg',\n",
       "  'image_id': 'b821cc12eede040e',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.520803, 0.940757, 0.603367, 0.988894], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/85ccc93d1f941931.jpg',\n",
       "  'image_id': '85ccc93d1f941931',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.369258, 0.746107, 0.152027, 1.0], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/15348e4f2c7ebe0f.jpg',\n",
       "  'image_id': '15348e4f2c7ebe0f',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.08732999999999999, 0.999973, 0.600616, 0.9999620000000001],\n",
       "           dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 0},\n",
       "   {'bbox': [array([0.09083300000000001, 1.0, 0.600928, 1.0], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 1}]},\n",
       " {'filename': 'images/822f20d881eebeb9.jpg',\n",
       "  'image_id': '822f20d881eebeb9',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.0, 1.0, 0.181437, 0.9631129999999999], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 0},\n",
       "   {'bbox': [array([0.0, 1.0, 0.170432, 0.9703510000000001], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 1}]},\n",
       " {'filename': 'images/eaa75967cbc70ac1.jpg',\n",
       "  'image_id': 'eaa75967cbc70ac1',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.09592200000000001, 0.8706450000000001, 0.12402300000000001,\n",
       "            0.57018], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/2ff2bb609a057f7a.jpg',\n",
       "  'image_id': '2ff2bb609a057f7a',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.182173, 0.798462, 0.273742, 0.942715], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/ddd3f96ff7f0bc78.jpg',\n",
       "  'image_id': 'ddd3f96ff7f0bc78',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.25188299999999997, 0.976808, 0.46091800000000005, 1.0],\n",
       "           dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 0}]},\n",
       " {'filename': 'images/d7d469cb4c7e8cd2.jpg',\n",
       "  'image_id': 'd7d469cb4c7e8cd2',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([8.1e-05, 0.999159, 0.435396, 1.0], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 0},\n",
       "   {'bbox': [array([0.0, 1.0, 0.43619399999999997, 1.0], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 1}]},\n",
       " {'filename': 'images/539e0871494dea5d.jpg',\n",
       "  'image_id': '539e0871494dea5d',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.20127799999999998, 0.815655, 0.0, 1.0], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/42298c9659ce6603.jpg',\n",
       "  'image_id': '42298c9659ce6603',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.301737, 0.792065, 0.19045299999999998, 0.83854], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/8e562fee6ff208f7.jpg',\n",
       "  'image_id': '8e562fee6ff208f7',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.013755000000000002, 0.5139520000000001, 0.47611800000000004,\n",
       "            0.99998], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 0},\n",
       "   {'bbox': [array([0.494666, 0.983858, 0.5461779999999999, 1.0], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 0}]},\n",
       " {'filename': 'images/cd727c5e0d2cfc1b.jpg',\n",
       "  'image_id': 'cd727c5e0d2cfc1b',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.048442, 0.9804959999999999, 0.427911, 0.806723], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 0}]},\n",
       " {'filename': 'images/f90eac5444f62b19.jpg',\n",
       "  'image_id': 'f90eac5444f62b19',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.405385, 0.8965549999999999, 0.624074, 1.0], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 3}]},\n",
       " {'filename': 'images/3f66fdf9688514d4.jpg',\n",
       "  'image_id': '3f66fdf9688514d4',\n",
       "  'height': 0,\n",
       "  'width': 0,\n",
       "  'annotations': [{'bbox': [array([0.050368, 0.9334950000000001, 0.250992, 1.0], dtype=object)],\n",
       "    'bbox_mode': 0,\n",
       "    'category_id': 0}]}]"
      ]
     },
     "execution_count": 224,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "# Create images labels setup in detectron2 style\n",
    "img_dicts = []\n",
    "for img in my_images:\n",
    "    record = {}\n",
    "    # Add image metadata\n",
    "    filename = \"images/\" + img + \".jpg\"\n",
    "    img_data = my_images_df[my_images_df[\"ImageID\"] == img].reset_index()\n",
    "    \n",
    "    record[\"file_name\"] = filename\n",
    "    record[\"image_id\"] = img\n",
    "    # TODO - use cv2\n",
    "    record[\"height\"] = 0\n",
    "    record[\"width\"] = 0\n",
    "    \n",
    "    # Create annotations list (contains labels of images)\n",
    "    annotations = []\n",
    "    for i in range(len(img_data)):\n",
    "        img_label = img_data.loc[i][\"ClassName\"]\n",
    "        category_id = img_data.loc[i][\"ClassID\"]\n",
    "        #print(f\"label: {img_label}\")\n",
    "        bbox = [img_data.loc[i][[\"XMin\", \"XMax\", \"YMin\", \"YMax\"]].values]\n",
    "        obj = {\n",
    "            \"bbox\": bbox,\n",
    "            \"bbox_mode\": 0, # TODO\n",
    "            #\"img_label\": img_label, # not needed\n",
    "            \"category_id\": category_id, \n",
    "             #\"segmentation\": [\"poly\"], # not needed for bounding boxes\n",
    "             #\"iscrowd\": 0 # not included as per detectron2 docs sure of this  \n",
    "        }\n",
    "        annotations.append(obj)\n",
    "        #print(annotations)\n",
    "    record[\"annotations\"] = annotations\n",
    "    img_dicts.append(record)\n",
    "img_dicts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "38"
      ]
     },
     "execution_count": 178,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(img_dicts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'file_path': 'images/d60e1c25e87d8b45.jpg',\n",
       " 'img_label': array(['Sink', 'Toilet'], dtype=object),\n",
       " 'bboxes': array([[0.      , 0.823265, 0.315739, 0.975746],\n",
       "        [0.405653, 1.      , 0.146613, 0.626838]])}"
      ]
     },
     "execution_count": 152,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "img_dicts[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "38"
      ]
     },
     "execution_count": 138,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(img_dicts)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# TK: Could I turn this into a script? Instead of it being in a notebook?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "ename": "TypeError",
     "evalue": "'_csv.reader' object is not subscriptable",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-60-71c69f9da2a3>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      3\u001b[0m     \u001b[0mreader\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcsv\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mreader\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfile\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 5\u001b[0;31m \u001b[0mreader\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;31mTypeError\u001b[0m: '_csv.reader' object is not subscriptable"
     ]
    }
   ],
   "source": [
    "import csv\n",
    "with open(\"validation-annotations-bbox.csv\") as file:\n",
    "    reader = csv.reader(file)   \n",
    "reader"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'Tortoise': '/m/011k07',\n",
       " 'Container': '/m/011q46kg',\n",
       " 'Magpie': '/m/012074',\n",
       " 'Sea turtle': '/m/0120dh',\n",
       " 'Football': '/m/01226z',\n",
       " 'Ambulance': '/m/012n7d',\n",
       " 'Ladder': '/m/012w5l',\n",
       " 'Toothbrush': '/m/012xff',\n",
       " 'Syringe': '/m/012ysf',\n",
       " 'Sink': '/m/0130jx',\n",
       " 'Toy': '/m/0138tl',\n",
       " 'Organ': '/m/013y1f',\n",
       " 'Cassette deck': '/m/01432t',\n",
       " 'Apple': '/m/014j1m',\n",
       " 'Human eye': '/m/014sv8',\n",
       " 'Cosmetics': '/m/014trl',\n",
       " 'Paddle': '/m/014y4n',\n",
       " 'Snowman': '/m/0152hh',\n",
       " 'Beer': '/m/01599',\n",
       " 'Chopsticks': '/m/01_5g',\n",
       " 'Human beard': '/m/015h_t',\n",
       " 'Bird': '/m/015p6',\n",
       " 'Parking meter': '/m/015qbp',\n",
       " 'Traffic light': '/m/015qff',\n",
       " 'Croissant': '/m/015wgc',\n",
       " 'Cucumber': '/m/015x4r',\n",
       " 'Radish': '/m/015x5n',\n",
       " 'Towel': '/m/0162_1',\n",
       " 'Doll': '/m/0167gd',\n",
       " 'Skull': '/m/016m2d',\n",
       " 'Washing machine': '/m/0174k2',\n",
       " 'Glove': '/m/0174n1',\n",
       " 'Tick': '/m/0175cv',\n",
       " 'Belt': '/m/0176mf',\n",
       " 'Sunglasses': '/m/017ftj',\n",
       " 'Banjo': '/m/018j2',\n",
       " 'Cart': '/m/018p4k',\n",
       " 'Ball': '/m/018xm',\n",
       " 'Backpack': '/m/01940j',\n",
       " 'Bicycle': '/m/0199g',\n",
       " 'Home appliance': '/m/019dx1',\n",
       " 'Centipede': '/m/019h78',\n",
       " 'Boat': '/m/019jd',\n",
       " 'Surfboard': '/m/019w40',\n",
       " 'Boot': '/m/01b638',\n",
       " 'Headphones': '/m/01b7fy',\n",
       " 'Hot dog': '/m/01b9xk',\n",
       " 'Shorts': '/m/01bfm9',\n",
       " 'Fast food': '/m/01_bhs',\n",
       " 'Bus': '/m/01bjv',\n",
       " 'Boy': '/m/01bl7v',\n",
       " 'Screwdriver': '/m/01bms0',\n",
       " 'Bicycle wheel': '/m/01bqk0',\n",
       " 'Barge': '/m/01btn',\n",
       " 'Laptop': '/m/01c648',\n",
       " 'Miniskirt': '/m/01cmb2',\n",
       " 'Drill': '/m/01d380',\n",
       " 'Dress': '/m/01d40f',\n",
       " 'Bear': '/m/01dws',\n",
       " 'Waffle': '/m/01dwsz',\n",
       " 'Pancake': '/m/01dwwc',\n",
       " 'Brown bear': '/m/01dxs',\n",
       " 'Woodpecker': '/m/01dy8n',\n",
       " 'Blue jay': '/m/01f8m5',\n",
       " 'Pretzel': '/m/01f91_',\n",
       " 'Bagel': '/m/01fb_0',\n",
       " 'Tower': '/m/01fdzj',\n",
       " 'Teapot': '/m/01fh4r',\n",
       " 'Person': '/m/01g317',\n",
       " 'Bow and arrow': '/m/01g3x7',\n",
       " 'Swimwear': '/m/01gkx_',\n",
       " 'Beehive': '/m/01gllr',\n",
       " 'Brassiere': '/m/01gmv2',\n",
       " 'Bee': '/m/01h3n',\n",
       " 'Bat': '/m/01h44',\n",
       " 'Starfish': '/m/01h8tj',\n",
       " 'Popcorn': '/m/01hrv5',\n",
       " 'Burrito': '/m/01j3zr',\n",
       " 'Chainsaw': '/m/01j4z9',\n",
       " 'Balloon': '/m/01j51',\n",
       " 'Wrench': '/m/01j5ks',\n",
       " 'Tent': '/m/01j61q',\n",
       " 'Vehicle registration plate': '/m/01jfm_',\n",
       " 'Lantern': '/m/01jfsr',\n",
       " 'Toaster': '/m/01k6s3',\n",
       " 'Flashlight': '/m/01kb5b',\n",
       " 'Billboard': '/m/01knjb',\n",
       " 'Tiara': '/m/01krhy',\n",
       " 'Limousine': '/m/01lcw4',\n",
       " 'Necklace': '/m/01llwg',\n",
       " 'Carnivore': '/m/01lrl',\n",
       " 'Scissors': '/m/01lsmm',\n",
       " 'Stairs': '/m/01lynh',\n",
       " 'Computer keyboard': '/m/01m2v',\n",
       " 'Printer': '/m/01m4t',\n",
       " 'Traffic sign': '/m/01mqdt',\n",
       " 'Chair': '/m/01mzpv',\n",
       " 'Shirt': '/m/01n4qj',\n",
       " 'Poster': '/m/01n5jq',\n",
       " 'Cheese': '/m/01nkt',\n",
       " 'Sock': '/m/01nq26',\n",
       " 'Fire hydrant': '/m/01pns0',\n",
       " 'Land vehicle': '/m/01prls',\n",
       " 'Earrings': '/m/01r546',\n",
       " 'Tie': '/m/01rkbr',\n",
       " 'Watercraft': '/m/01rzcn',\n",
       " 'Cabinetry': '/m/01s105',\n",
       " 'Suitcase': '/m/01s55n',\n",
       " 'Muffin': '/m/01tcjp',\n",
       " 'Bidet': '/m/01vbnl',\n",
       " 'Snack': '/m/01ww8y',\n",
       " 'Snowmobile': '/m/01x3jk',\n",
       " 'Clock': '/m/01x3z',\n",
       " 'Medical equipment': '/m/01xgg_',\n",
       " 'Cattle': '/m/01xq0k1',\n",
       " 'Cello': '/m/01xqw',\n",
       " 'Jet ski': '/m/01xs3r',\n",
       " 'Camel': '/m/01x_v',\n",
       " 'Coat': '/m/01xygc',\n",
       " 'Suit': '/m/01xyhv',\n",
       " 'Desk': '/m/01y9k5',\n",
       " 'Cat': '/m/01yrx',\n",
       " 'Bronze sculpture': '/m/01yx86',\n",
       " 'Juice': '/m/01z1kdw',\n",
       " 'Gondola': '/m/02068x',\n",
       " 'Beetle': '/m/020jm',\n",
       " 'Cannon': '/m/020kz',\n",
       " 'Computer mouse': '/m/020lf',\n",
       " 'Cookie': '/m/021mn',\n",
       " 'Office building': '/m/021sj1',\n",
       " 'Fountain': '/m/0220r2',\n",
       " 'Coin': '/m/0242l',\n",
       " 'Calculator': '/m/024d2',\n",
       " 'Cocktail': '/m/024g6',\n",
       " 'Computer monitor': '/m/02522',\n",
       " 'Box': '/m/025dyy',\n",
       " 'Stapler': '/m/025fsf',\n",
       " 'Christmas tree': '/m/025nd',\n",
       " 'Cowboy hat': '/m/025rp__',\n",
       " 'Hiking equipment': '/m/0268lbt',\n",
       " 'Studio couch': '/m/026qbn5',\n",
       " 'Drum': '/m/026t6',\n",
       " 'Dessert': '/m/0270h',\n",
       " 'Wine rack': '/m/0271qf7',\n",
       " 'Drink': '/m/0271t',\n",
       " 'Zucchini': '/m/027pcv',\n",
       " 'Ladle': '/m/027rl48',\n",
       " 'Human mouth': '/m/0283dt1',\n",
       " 'Dairy': '/m/0284d',\n",
       " 'Dice': '/m/029b3',\n",
       " 'Oven': '/m/029bxz',\n",
       " 'Dinosaur': '/m/029tx',\n",
       " 'Ratchet': '/m/02bm9n',\n",
       " 'Couch': '/m/02crq1',\n",
       " 'Cricket ball': '/m/02ctlc',\n",
       " 'Winter melon': '/m/02cvgx',\n",
       " 'Spatula': '/m/02d1br',\n",
       " 'Whiteboard': '/m/02d9qx',\n",
       " 'Pencil sharpener': '/m/02ddwp',\n",
       " 'Door': '/m/02dgv',\n",
       " 'Hat': '/m/02dl1y',\n",
       " 'Shower': '/m/02f9f_',\n",
       " 'Eraser': '/m/02fh7f',\n",
       " 'Fedora': '/m/02fq_6',\n",
       " 'Guacamole': '/m/02g30s',\n",
       " 'Dagger': '/m/02gzp',\n",
       " 'Scarf': '/m/02h19r',\n",
       " 'Dolphin': '/m/02hj4',\n",
       " 'Sombrero': '/m/02jfl0',\n",
       " 'Tin can': '/m/02jnhm',\n",
       " 'Mug': '/m/02jvh9',\n",
       " 'Tap': '/m/02jz0l',\n",
       " 'Harbor seal': '/m/02l8p9',\n",
       " 'Stretcher': '/m/02lbcq',\n",
       " 'Can opener': '/m/02mqfb',\n",
       " 'Goggles': '/m/02_n6y',\n",
       " 'Human body': '/m/02p0tk3',\n",
       " 'Roller skates': '/m/02p3w7d',\n",
       " 'Coffee cup': '/m/02p5f1q',\n",
       " 'Cutting board': '/m/02pdsw',\n",
       " 'Blender': '/m/02pjr4',\n",
       " 'Plumbing fixture': '/m/02pkr5',\n",
       " 'Stop sign': '/m/02pv19',\n",
       " 'Office supplies': '/m/02rdsp',\n",
       " 'Volleyball': '/m/02rgn06',\n",
       " 'Vase': '/m/02s195',\n",
       " 'Slow cooker': '/m/02tsc9',\n",
       " 'Wardrobe': '/m/02vkqh8',\n",
       " 'Coffee': '/m/02vqfm',\n",
       " 'Whisk': '/m/02vwcm',\n",
       " 'Paper towel': '/m/02w3r3',\n",
       " 'Personal care': '/m/02w3_ws',\n",
       " 'Food': '/m/02wbm',\n",
       " 'Sun hat': '/m/02wbtzl',\n",
       " 'Tree house': '/m/02wg_p',\n",
       " 'Flying disc': '/m/02wmf',\n",
       " 'Skirt': '/m/02wv6h6',\n",
       " 'Gas stove': '/m/02wv84t',\n",
       " 'Salt and pepper shakers': '/m/02x8cch',\n",
       " 'Mechanical fan': '/m/02x984l',\n",
       " 'Face powder': '/m/02xb7qb',\n",
       " 'Fax': '/m/02xqq',\n",
       " 'Fruit': '/m/02xwb',\n",
       " 'French fries': '/m/02y6n',\n",
       " 'Nightstand': '/m/02z51p',\n",
       " 'Barrel': '/m/02zn6n',\n",
       " 'Kite': '/m/02zt3',\n",
       " 'Tart': '/m/02zvsm',\n",
       " 'Treadmill': '/m/030610',\n",
       " 'Fox': '/m/0306r',\n",
       " 'Flag': '/m/03120',\n",
       " 'Horn': '/m/0319l',\n",
       " 'Window blind': '/m/031b6r',\n",
       " 'Human foot': '/m/031n1',\n",
       " 'Golf cart': '/m/0323sq',\n",
       " 'Jacket': '/m/032b3c',\n",
       " 'Egg': '/m/033cnk',\n",
       " 'Street light': '/m/033rq4',\n",
       " 'Guitar': '/m/0342h',\n",
       " 'Pillow': '/m/034c16',\n",
       " 'Human leg': '/m/035r7c',\n",
       " 'Isopod': '/m/035vxb',\n",
       " 'Grape': '/m/0388q',\n",
       " 'Human ear': '/m/039xj_',\n",
       " 'Power plugs and sockets': '/m/03bbps',\n",
       " 'Panda': '/m/03bj1',\n",
       " 'Giraffe': '/m/03bk1',\n",
       " 'Woman': '/m/03bt1vf',\n",
       " 'Door handle': '/m/03c7gz',\n",
       " 'Rhinoceros': '/m/03d443',\n",
       " 'Bathtub': '/m/03dnzn',\n",
       " 'Goldfish': '/m/03fj2',\n",
       " 'Houseplant': '/m/03fp41',\n",
       " 'Goat': '/m/03fwl',\n",
       " 'Baseball bat': '/m/03g8mr',\n",
       " 'Baseball glove': '/m/03grzl',\n",
       " 'Mixing bowl': '/m/03hj559',\n",
       " 'Marine invertebrates': '/m/03hl4l9',\n",
       " 'Kitchen utensil': '/m/03hlz0c',\n",
       " 'Light switch': '/m/03jbxj',\n",
       " 'House': '/m/03jm5',\n",
       " 'Horse': '/m/03k3r',\n",
       " 'Stationary bicycle': '/m/03kt2w',\n",
       " 'Hammer': '/m/03l9g',\n",
       " 'Ceiling fan': '/m/03ldnb',\n",
       " 'Sofa bed': '/m/03m3pdh',\n",
       " 'Adhesive tape': '/m/03m3vtv',\n",
       " 'Harp': '/m/03m5k',\n",
       " 'Sandal': '/m/03nfch',\n",
       " 'Bicycle helmet': '/m/03p3bw',\n",
       " 'Saucer': '/m/03q5c7',\n",
       " 'Harpsichord': '/m/03q5t',\n",
       " 'Human hair': '/m/03q69',\n",
       " 'Heater': '/m/03qhv5',\n",
       " 'Harmonica': '/m/03qjg',\n",
       " 'Hamster': '/m/03qrc',\n",
       " 'Curtain': '/m/03rszm',\n",
       " 'Bed': '/m/03ssj5',\n",
       " 'Kettle': '/m/03s_tn',\n",
       " 'Fireplace': '/m/03tw93',\n",
       " 'Scale': '/m/03txqz',\n",
       " 'Drinking straw': '/m/03v5tg',\n",
       " 'Insect': '/m/03vt0',\n",
       " 'Hair dryer': '/m/03wvsk',\n",
       " 'Kitchenware': '/m/03_wxk',\n",
       " 'Indoor rower': '/m/03wym',\n",
       " 'Invertebrate': '/m/03xxp',\n",
       " 'Food processor': '/m/03y6mg',\n",
       " 'Bookcase': '/m/03__z0',\n",
       " 'Refrigerator': '/m/040b_t',\n",
       " 'Wood-burning stove': '/m/04169hn',\n",
       " 'Punching bag': '/m/0420v5',\n",
       " 'Common fig': '/m/043nyj',\n",
       " 'Cocktail shaker': '/m/0440zs',\n",
       " 'Jaguar': '/m/0449p',\n",
       " 'Golf ball': '/m/044r5d',\n",
       " 'Fashion accessory': '/m/0463sg',\n",
       " 'Alarm clock': '/m/046dlr',\n",
       " 'Filing cabinet': '/m/047j0r',\n",
       " 'Artichoke': '/m/047v4b',\n",
       " 'Table': '/m/04bcr3',\n",
       " 'Tableware': '/m/04brg2',\n",
       " 'Kangaroo': '/m/04c0y',\n",
       " 'Koala': '/m/04cp_',\n",
       " 'Knife': '/m/04ctx',\n",
       " 'Bottle': '/m/04dr76w',\n",
       " 'Bottle opener': '/m/04f5ws',\n",
       " 'Lynx': '/m/04g2r',\n",
       " 'Lavender': '/m/04gth',\n",
       " 'Lighthouse': '/m/04h7h',\n",
       " 'Dumbbell': '/m/04h8sr',\n",
       " 'Human head': '/m/04hgtk',\n",
       " 'Bowl': '/m/04kkgm',\n",
       " 'Humidifier': '/m/04lvq_',\n",
       " 'Porch': '/m/04m6gz',\n",
       " 'Lizard': '/m/04m9y',\n",
       " 'Billiard table': '/m/04p0qw',\n",
       " 'Mammal': '/m/04rky',\n",
       " 'Mouse': '/m/04rmv',\n",
       " 'Motorcycle': '/m/04_sv',\n",
       " 'Musical instrument': '/m/04szw',\n",
       " 'Swim cap': '/m/04tn4x',\n",
       " 'Frying pan': '/m/04v6l4',\n",
       " 'Snowplow': '/m/04vv5k',\n",
       " 'Bathroom cabinet': '/m/04y4h8h',\n",
       " 'Missile': '/m/04ylt',\n",
       " 'Bust': '/m/04yqq2',\n",
       " 'Man': '/m/04yx4',\n",
       " 'Waffle iron': '/m/04z4wx',\n",
       " 'Milk': '/m/04zpv',\n",
       " 'Ring binder': '/m/04zwwv',\n",
       " 'Plate': '/m/050gv4',\n",
       " 'Mobile phone': '/m/050k8',\n",
       " 'Baked goods': '/m/052lwg6',\n",
       " 'Mushroom': '/m/052sf',\n",
       " 'Crutch': '/m/05441v',\n",
       " 'Pitcher': '/m/054fyh',\n",
       " 'Mirror': '/m/054_l',\n",
       " 'Lifejacket': '/m/054xkw',\n",
       " 'Table tennis racket': '/m/05_5p_0',\n",
       " 'Pencil case': '/m/05676x',\n",
       " 'Musical keyboard': '/m/057cc',\n",
       " 'Scoreboard': '/m/057p5t',\n",
       " 'Briefcase': '/m/0584n8',\n",
       " 'Kitchen knife': '/m/058qzx',\n",
       " 'Nail': '/m/05bm6',\n",
       " 'Tennis ball': '/m/05ctyq',\n",
       " 'Plastic bag': '/m/05gqfk',\n",
       " 'Oboe': '/m/05kms',\n",
       " 'Chest of drawers': '/m/05kyg_',\n",
       " 'Ostrich': '/m/05n4y',\n",
       " 'Piano': '/m/05r5c',\n",
       " 'Girl': '/m/05r655',\n",
       " 'Plant': '/m/05s2s',\n",
       " 'Potato': '/m/05vtc',\n",
       " 'Hair spray': '/m/05w9t9',\n",
       " 'Sports equipment': '/m/05y5lj',\n",
       " 'Pasta': '/m/05z55',\n",
       " 'Penguin': '/m/05z6w',\n",
       " 'Pumpkin': '/m/05zsy',\n",
       " 'Pear': '/m/061_f',\n",
       " 'Infant bed': '/m/061hd_',\n",
       " 'Polar bear': '/m/0633h',\n",
       " 'Mixer': '/m/063rgb',\n",
       " 'Cupboard': '/m/0642b4',\n",
       " 'Jacuzzi': '/m/065h6l',\n",
       " 'Pizza': '/m/0663v',\n",
       " 'Digital clock': '/m/06_72j',\n",
       " 'Pig': '/m/068zj',\n",
       " 'Reptile': '/m/06bt6',\n",
       " 'Rifle': '/m/06c54',\n",
       " 'Lipstick': '/m/06c7f7',\n",
       " 'Skateboard': '/m/06_fw',\n",
       " 'Raven': '/m/06j2d',\n",
       " 'High heels': '/m/06k2mb',\n",
       " 'Red panda': '/m/06l9r',\n",
       " 'Rose': '/m/06m11',\n",
       " 'Rabbit': '/m/06mf6',\n",
       " 'Sculpture': '/m/06msq',\n",
       " 'Saxophone': '/m/06ncr',\n",
       " 'Shotgun': '/m/06nrc',\n",
       " 'Seafood': '/m/06nwz',\n",
       " 'Submarine sandwich': '/m/06pcq',\n",
       " 'Snowboard': '/m/06__v',\n",
       " 'Sword': '/m/06y5r',\n",
       " 'Picture frame': '/m/06z37_',\n",
       " 'Sushi': '/m/07030',\n",
       " 'Loveseat': '/m/0703r8',\n",
       " 'Ski': '/m/071p9',\n",
       " 'Squirrel': '/m/071qp',\n",
       " 'Tripod': '/m/073bxn',\n",
       " 'Stethoscope': '/m/073g6',\n",
       " 'Submarine': '/m/074d1',\n",
       " 'Scorpion': '/m/0755b',\n",
       " 'Segway': '/m/076bq',\n",
       " 'Training bench': '/m/076lb9',\n",
       " 'Snake': '/m/078jl',\n",
       " 'Coffee table': '/m/078n6m',\n",
       " 'Skyscraper': '/m/079cl',\n",
       " 'Sheep': '/m/07bgp',\n",
       " 'Television': '/m/07c52',\n",
       " 'Trombone': '/m/07c6l',\n",
       " 'Tea': '/m/07clx',\n",
       " 'Tank': '/m/07cmd',\n",
       " 'Taco': '/m/07crc',\n",
       " 'Telephone': '/m/07cx4',\n",
       " 'Torch': '/m/07dd4',\n",
       " 'Tiger': '/m/07dm6',\n",
       " 'Strawberry': '/m/07fbm7',\n",
       " 'Trumpet': '/m/07gql',\n",
       " 'Tree': '/m/07j7r',\n",
       " 'Tomato': '/m/07j87',\n",
       " 'Train': '/m/07jdr',\n",
       " 'Tool': '/m/07k1x',\n",
       " 'Picnic basket': '/m/07kng9',\n",
       " 'Cooking spray': '/m/07mcwg',\n",
       " 'Trousers': '/m/07mhn',\n",
       " 'Bowling equipment': '/m/07pj7bq',\n",
       " 'Football helmet': '/m/07qxg_',\n",
       " 'Truck': '/m/07r04',\n",
       " 'Measuring cup': '/m/07v9_z',\n",
       " 'Coffeemaker': '/m/07xyvk',\n",
       " 'Violin': '/m/07y_7',\n",
       " 'Vehicle': '/m/07yv9',\n",
       " 'Handbag': '/m/080hkjn',\n",
       " 'Paper cutter': '/m/080n7g',\n",
       " 'Wine': '/m/081qc',\n",
       " 'Weapon': '/m/083kb',\n",
       " 'Wheel': '/m/083wq',\n",
       " 'Worm': '/m/084hf',\n",
       " 'Wok': '/m/084rd',\n",
       " 'Whale': '/m/084zz',\n",
       " 'Zebra': '/m/0898b',\n",
       " 'Auto part': '/m/08dz3q',\n",
       " 'Jug': '/m/08hvt4',\n",
       " 'Pizza cutter': '/m/08ks85',\n",
       " 'Cream': '/m/08p92x',\n",
       " 'Monkey': '/m/08pbxl',\n",
       " 'Lion': '/m/096mb',\n",
       " 'Bread': '/m/09728',\n",
       " 'Platter': '/m/099ssp',\n",
       " 'Chicken': '/m/09b5t',\n",
       " 'Eagle': '/m/09csl',\n",
       " 'Helicopter': '/m/09ct_',\n",
       " 'Owl': '/m/09d5_',\n",
       " 'Duck': '/m/09ddx',\n",
       " 'Turtle': '/m/09dzg',\n",
       " 'Hippopotamus': '/m/09f20',\n",
       " 'Crocodile': '/m/09f_2',\n",
       " 'Toilet': '/m/09g1w',\n",
       " 'Toilet paper': '/m/09gtd',\n",
       " 'Squid': '/m/09gys',\n",
       " 'Clothing': '/m/09j2d',\n",
       " 'Footwear': '/m/09j5n',\n",
       " 'Lemon': '/m/09k_b',\n",
       " 'Spider': '/m/09kmb',\n",
       " 'Deer': '/m/09kx5',\n",
       " 'Frog': '/m/09ld4',\n",
       " 'Banana': '/m/09qck',\n",
       " 'Rocket': '/m/09rvcxw',\n",
       " 'Wine glass': '/m/09tvcd',\n",
       " 'Countertop': '/m/0b3fp9',\n",
       " 'Tablet computer': '/m/0bh9flk',\n",
       " 'Waste container': '/m/0bjyj5',\n",
       " 'Swimming pool': '/m/0b_rs',\n",
       " 'Dog': '/m/0bt9lr',\n",
       " 'Book': '/m/0bt_c3',\n",
       " 'Elephant': '/m/0bwd_0j',\n",
       " 'Shark': '/m/0by6g',\n",
       " 'Candle': '/m/0c06p',\n",
       " 'Leopard': '/m/0c29q',\n",
       " 'Axe': '/m/0c2jj',\n",
       " 'Hand dryer': '/m/0c3m8g',\n",
       " 'Soap dispenser': '/m/0c3mkw',\n",
       " 'Porcupine': '/m/0c568',\n",
       " 'Flower': '/m/0c9ph5',\n",
       " 'Canary': '/m/0ccs93',\n",
       " 'Cheetah': '/m/0cd4d',\n",
       " 'Palm tree': '/m/0cdl1',\n",
       " 'Hamburger': '/m/0cdn1',\n",
       " 'Maple': '/m/0cffdh',\n",
       " 'Building': '/m/0cgh4',\n",
       " 'Fish': '/m/0ch_cf',\n",
       " 'Lobster': '/m/0cjq5',\n",
       " 'Asparagus': '/m/0cjs7',\n",
       " 'Furniture': '/m/0c_jw',\n",
       " 'Hedgehog': '/m/0cl4p',\n",
       " 'Airplane': '/m/0cmf2',\n",
       " 'Spoon': '/m/0cmx8',\n",
       " 'Otter': '/m/0cn6p',\n",
       " 'Bull': '/m/0cnyhnx',\n",
       " 'Oyster': '/m/0_cp5',\n",
       " 'Horizontal bar': '/m/0cqn2',\n",
       " 'Convenience store': '/m/0crjs',\n",
       " 'Bomb': '/m/0ct4f',\n",
       " 'Bench': '/m/0cvnqh',\n",
       " 'Ice cream': '/m/0cxn2',\n",
       " 'Caterpillar': '/m/0cydv',\n",
       " 'Butterfly': '/m/0cyf8',\n",
       " 'Parachute': '/m/0cyfs',\n",
       " 'Orange': '/m/0cyhj_',\n",
       " 'Antelope': '/m/0czz2',\n",
       " 'Beaker': '/m/0d20w4',\n",
       " 'Moths and butterflies': '/m/0d_2m',\n",
       " 'Window': '/m/0d4v4',\n",
       " 'Closet': '/m/0d4w1',\n",
       " 'Castle': '/m/0d5gx',\n",
       " 'Jellyfish': '/m/0d8zb',\n",
       " 'Goose': '/m/0dbvp',\n",
       " 'Mule': '/m/0dbzx',\n",
       " 'Swan': '/m/0dftk',\n",
       " 'Peach': '/m/0dj6p',\n",
       " 'Coconut': '/m/0djtd',\n",
       " 'Seat belt': '/m/0dkzw',\n",
       " 'Raccoon': '/m/0dq75',\n",
       " 'Chisel': '/m/0_dqb',\n",
       " 'Fork': '/m/0dt3t',\n",
       " 'Lamp': '/m/0dtln',\n",
       " 'Camera': '/m/0dv5r',\n",
       " 'Squash': '/m/0dv77',\n",
       " 'Racket': '/m/0dv9c',\n",
       " 'Human face': '/m/0dzct',\n",
       " 'Human arm': '/m/0dzf4',\n",
       " 'Vegetable': '/m/0f4s2w',\n",
       " 'Diaper': '/m/0f571',\n",
       " 'Unicycle': '/m/0f6nr',\n",
       " 'Falcon': '/m/0f6wt',\n",
       " 'Chime': '/m/0f8s22',\n",
       " 'Snail': '/m/0f9_l',\n",
       " 'Shellfish': '/m/0fbdv',\n",
       " 'Cabbage': '/m/0fbw6',\n",
       " 'Carrot': '/m/0fj52s',\n",
       " 'Mango': '/m/0fldg',\n",
       " 'Jeans': '/m/0fly7',\n",
       " 'Flowerpot': '/m/0fm3zh',\n",
       " 'Pineapple': '/m/0fp6w',\n",
       " 'Drawer': '/m/0fqfqc',\n",
       " 'Stool': '/m/0fqt361',\n",
       " 'Envelope': '/m/0frqm',\n",
       " 'Cake': '/m/0fszt',\n",
       " 'Dragonfly': '/m/0ft9s',\n",
       " 'Sunflower': '/m/0ftb8',\n",
       " 'Microwave oven': '/m/0fx9l',\n",
       " 'Honeycomb': '/m/0fz0h',\n",
       " 'Marine mammal': '/m/0gd2v',\n",
       " 'Sea lion': '/m/0gd36',\n",
       " 'Ladybug': '/m/0gj37',\n",
       " 'Shelf': '/m/0gjbg72',\n",
       " 'Watch': '/m/0gjkl',\n",
       " 'Candy': '/m/0gm28',\n",
       " 'Salad': '/m/0grw1',\n",
       " 'Parrot': '/m/0gv1x',\n",
       " 'Handgun': '/m/0gxl3',\n",
       " 'Sparrow': '/m/0h23m',\n",
       " 'Van': '/m/0h2r6',\n",
       " 'Grinder': '/m/0h8jyh6',\n",
       " 'Spice rack': '/m/0h8kx63',\n",
       " 'Light bulb': '/m/0h8l4fh',\n",
       " 'Corded phone': '/m/0h8lkj8',\n",
       " 'Sports uniform': '/m/0h8mhzd',\n",
       " 'Tennis racket': '/m/0h8my_4',\n",
       " 'Wall clock': '/m/0h8mzrc',\n",
       " 'Serving tray': '/m/0h8n27j',\n",
       " 'Kitchen & dining room table': '/m/0h8n5zk',\n",
       " 'Dog bed': '/m/0h8n6f9',\n",
       " 'Cake stand': '/m/0h8n6ft',\n",
       " 'Cat furniture': '/m/0h8nm9j',\n",
       " 'Bathroom accessory': '/m/0h8nr_l',\n",
       " 'Facial tissue holder': '/m/0h8nsvg',\n",
       " 'Pressure cooker': '/m/0h8ntjv',\n",
       " 'Kitchen appliance': '/m/0h99cwc',\n",
       " 'Tire': '/m/0h9mv',\n",
       " 'Ruler': '/m/0hdln',\n",
       " 'Luggage and bags': '/m/0hf58v5',\n",
       " 'Microphone': '/m/0hg7b',\n",
       " 'Broccoli': '/m/0hkxq',\n",
       " 'Umbrella': '/m/0hnnb',\n",
       " 'Pastry': '/m/0hnyx',\n",
       " 'Grapefruit': '/m/0hqkz',\n",
       " 'Band-aid': '/m/0j496',\n",
       " 'Animal': '/m/0jbk',\n",
       " 'Bell pepper': '/m/0jg57',\n",
       " 'Turkey': '/m/0jly1',\n",
       " 'Lily': '/m/0jqgx',\n",
       " 'Pomegranate': '/m/0jwn_',\n",
       " 'Doughnut': '/m/0jy4k',\n",
       " 'Glasses': '/m/0jyfg',\n",
       " 'Human nose': '/m/0k0pj',\n",
       " 'Pen': '/m/0k1tl',\n",
       " 'Ant': '/m/0_k2',\n",
       " 'Car': '/m/0k4j',\n",
       " 'Aircraft': '/m/0k5j',\n",
       " 'Human hand': '/m/0k65p',\n",
       " 'Skunk': '/m/0km7z',\n",
       " 'Teddy bear': '/m/0kmg4',\n",
       " 'Watermelon': '/m/0kpqd',\n",
       " 'Cantaloupe': '/m/0kpt_',\n",
       " 'Dishwasher': '/m/0ky7b',\n",
       " 'Flute': '/m/0l14j_',\n",
       " 'Balance beam': '/m/0l3ms',\n",
       " 'Sandwich': '/m/0l515',\n",
       " 'Shrimp': '/m/0ll1f78',\n",
       " 'Sewing machine': '/m/0llzx',\n",
       " 'Binoculars': '/m/0lt4_',\n",
       " 'Rays and skates': '/m/0m53l',\n",
       " 'Ipod': '/m/0mcx2',\n",
       " 'Accordion': '/m/0mkg',\n",
       " 'Willow': '/m/0mw_6',\n",
       " 'Crab': '/m/0n28_',\n",
       " 'Crown': '/m/0nl46',\n",
       " 'Seahorse': '/m/0nybt',\n",
       " 'Perfume': '/m/0p833',\n",
       " 'Alpaca': '/m/0pcr',\n",
       " 'Taxi': '/m/0pg52',\n",
       " 'Canoe': '/m/0ph39',\n",
       " 'Remote control': '/m/0qjjc',\n",
       " 'Wheelchair': '/m/0qmmr',\n",
       " 'Rugby ball': '/m/0wdt60w',\n",
       " 'Armadillo': '/m/0xfy',\n",
       " 'Maracas': '/m/0xzly',\n",
       " 'Helmet': '/m/0zvk5'}"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import csv\n",
    "with open(\"class-descriptions-boxable.csv\") as file:\n",
    "    reader = csv.reader(file)\n",
    "    dict_list = {row[1]:row[0] for row in reader}\n",
    "dict_list"
   ]
  }
 ],
 "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}