prediction_tutorial.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# nuScenes prediction tutorial\n",
    "<img src=\"trajectory.gif\" width=\"300\" align=\"center\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This notebook serves as an introduction to the new functionality added to the nuScenes devkit for the prediction challenge.\n",
    "\n",
    "It is organized into the following sections:\n",
    "\n",
    "1. Data splits for the challenge\n",
    "2. Getting past and future data for an agent \n",
    "3. Changes to the Map API\n",
    "4. Overview of input representation\n",
    "5. Model implementations\n",
    "6. Making a submission to the challenge"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "======\n",
      "Loading NuScenes tables for version v1.0-mini...\n",
      "23 category,\n",
      "8 attribute,\n",
      "4 visibility,\n",
      "911 instance,\n",
      "12 sensor,\n",
      "120 calibrated_sensor,\n",
      "31206 ego_pose,\n",
      "8 log,\n",
      "10 scene,\n",
      "404 sample,\n",
      "31206 sample_data,\n",
      "18538 sample_annotation,\n",
      "4 map,\n",
      "Done loading in 0.8 seconds.\n",
      "======\n",
      "Reverse indexing ...\n",
      "Done reverse indexing in 0.2 seconds.\n",
      "======\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "'./data/sets/nuscenes'"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from nuscenes import NuScenes\n",
    "\n",
    "# This is the path where you stored your copy of the nuScenes dataset.\n",
    "DATAROOT = './data/sets/nuscenes'\n",
    "\n",
    "nusc = NuScenes('v1.0-mini', dataroot=DATAROOT)\n",
    "nusc.dataroot"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Data Splits for the Prediction Challenge\n",
    "\n",
    "This section assumes basic familiarity with the nuScenes [schema](https://www.nuscenes.org/data-format?externalData=all&mapData=all&modalities=Any).\n",
    "\n",
    "The goal of the nuScenes prediction challenge is to predict the future location of agents in the nuScenes dataset. Agents are indexed by an instance token and a sample token. To get a list of agents in the train and val split of the challenge, we provide a function called `get_prediction_challenge_split`.\n",
    "\n",
    "The get_prediction_challenge_split function returns a list of strings of the form {instance_token}_{sample_token}. In the next section, we show how to use an instance token and sample token to query data for the prediction challenge."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['bc38961ca0ac4b14ab90e547ba79fbb6_39586f9d59004284a7114a68825e8eec',\n",
       " 'bc38961ca0ac4b14ab90e547ba79fbb6_356d81f38dd9473ba590f39e266f54e5',\n",
       " 'bc38961ca0ac4b14ab90e547ba79fbb6_e0845f5322254dafadbbed75aaa07969',\n",
       " 'bc38961ca0ac4b14ab90e547ba79fbb6_c923fe08b2ff4e27975d2bf30934383b',\n",
       " 'bc38961ca0ac4b14ab90e547ba79fbb6_f1e3d9d08f044c439ce86a2d6fcca57b']"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from nuscenes.eval.prediction.splits import get_prediction_challenge_split\n",
    "from nuscenes.eval.prediction.config import load_prediction_config\n",
    "from nuscenes.eval.prediction.splits import get_prediction_challenge_split\n",
    "from nuscenes.prediction import PredictHelper\n",
    "from nuscenes.prediction.models.physics import ConstantVelocityHeading, PhysicsOracle\n",
    "helper = PredictHelper(nusc)\n",
    "mini_train = get_prediction_challenge_split('mini_train',dataroot=DATAROOT)\n",
    "mini_train[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "742"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(mini_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Getting past and future data for an agent\n",
    "\n",
    "We provide a class called `PredictHelper` that provides methods for querying past and future data for an agent. This class is instantiated by wrapping an instance of the `NuScenes` class. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "from nuscenes.prediction import PredictHelper\n",
    "helper = PredictHelper(nusc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To get the data for an agent at a particular point in time, use the `get_sample_annotation` method."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'annotation' 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-10-0a77a1d01128>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mannotation\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'num_lidar_pts'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m: name 'annotation' is not defined"
     ]
    }
   ],
   "source": [
    "annotation['num_lidar_pts']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "from nuscenes.utils.data_classes import LidarPointCloud, RadarPointCloud, Box\n",
    "def get_lidar_data_pc(lidar_data_token, nsweeps=1):\n",
    "    sd_record = nusc.get('sample_data', lidar_data_token)\n",
    "    sample_rec = nusc.get('sample', sd_record['sample_token'])\n",
    "    chan = sd_record['channel']\n",
    "    ref_channel = 'LIDAR_TOP'\n",
    "    ref_sd_token = sample_rec['data'][ref_channel]\n",
    "    ref_sd_record = nusc.get('sample_data', ref_sd_token)\n",
    "    pc, times = LidarPointCloud.from_file_multisweep(nusc, sample_rec, chan, ref_channel, nsweeps=nsweeps)\n",
    "    return pc, times"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'sample_token' 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-12-37ece0542594>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      2\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mnuscenes\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mutils\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mgeometry_utils\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 4\u001b[0;31m \u001b[0msample\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnusc\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'sample'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msample_token\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      5\u001b[0m \u001b[0mlidar_token\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msample\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'data'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'LIDAR_TOP'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      6\u001b[0m \u001b[0mselected_anntokens\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mannotation\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'token'\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 'sample_token' is not defined"
     ]
    }
   ],
   "source": [
    "from nuscenes.utils.geometry_utils import view_points, box_in_image, BoxVisibility, transform_matrix\n",
    "from nuscenes.utils.geometry_utils import *\n",
    "\n",
    "sample = nusc.get('sample', sample_token)\n",
    "lidar_token = sample['data']['LIDAR_TOP']\n",
    "selected_anntokens = annotation['token']\n",
    "\n",
    "nusc.render_annotation(selected_anntokens)\n",
    "\n",
    "ann_record = nusc.get('sample_annotation', selected_anntokens)\n",
    "sample_record = nusc.get('sample', ann_record['sample_token'])\n",
    "\n",
    "# Figure out which camera the object is fully visible in (this may return nothing)\n",
    "boxes, cam = [], []\n",
    "cam = 'CAM_FRONT'\n",
    "_, boxes, _ = nusc.get_sample_data(sample_record['data'][cam], box_vis_level=BoxVisibility.ANY,\n",
    "                                        selected_anntokens=[selected_anntokens])\n",
    "cam = sample_record['data'][cam]\n",
    "lidar = sample_record['data']['LIDAR_TOP']\n",
    "pc, times = get_lidar_data_pc(lidar, nsweeps=5)\n",
    "print(pc.shape)\n",
    "_, boxes, _ = nusc.get_sample_data(lidar, selected_anntokens=[selected_anntokens])\n",
    "mask = points_in_box(boxes[0], pc.points[:3,:])\n",
    "\n",
    "# print(len(boxes))\n",
    "# print(data_path)\n",
    "# print(pc.points[:,mask].shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def project_points_to_image(pc, lidar_data, cam):\n",
    "    #assumes lidar data in sensor frame\n",
    "    cs_record = nusc.get('calibrated_sensor', lidar_data['calibrated_sensor_token'])\n",
    "    pc.rotate(Quaternion(cs_record['rotation']).rotation_matrix)\n",
    "    pc.translate(np.array(cs_record['translation']))\n",
    "\n",
    "    # Second step: transform to the global frame.\n",
    "    poserecord = nusc.get('ego_pose', lidar_data['ego_pose_token'])\n",
    "    pc.rotate(Quaternion(poserecord['rotation']).rotation_matrix)\n",
    "    pc.translate(np.array(poserecord['translation']))\n",
    "    \n",
    "    global_frame_pc = copy.deepcopy(pc)  \n",
    "\n",
    "    # Third step: transform into the ego vehicle frame for the timestamp of the image.\n",
    "    poserecord = nusc.get('ego_pose', cam['ego_pose_token'])\n",
    "    pc.translate(-np.array(poserecord['translation']))\n",
    "    pc.rotate(Quaternion(poserecord['rotation']).rotation_matrix.T)\n",
    "\n",
    "    # Fourth step: transform into the camera.\n",
    "    cs_record = nusc.get('calibrated_sensor', cam['calibrated_sensor_token'])\n",
    "    pc.translate(-np.array(cs_record['translation']))\n",
    "    pc.rotate(Quaternion(cs_record['rotation']).rotation_matrix.T)\n",
    "    \n",
    "    return pc, global_frame_pc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "4 4\n",
      "1\n",
      "0 0\n",
      "4 4\n",
      "2\n",
      "10 10\n",
      "0 0\n",
      "4 4\n",
      "3\n",
      "13 13\n",
      "10 10\n",
      "0 0\n",
      "4 4\n",
      "4\n",
      "18 18\n",
      "13 13\n",
      "10 10\n",
      "0 0\n",
      "4 4\n",
      "5\n",
      "20 20\n",
      "18 18\n",
      "13 13\n",
      "10 10\n",
      "0 0\n",
      "4 4\n",
      "6\n",
      "24 24\n",
      "20 20\n",
      "18 18\n",
      "13 13\n",
      "10 10\n",
      "0 0\n",
      "7\n",
      "30 30\n",
      "24 24\n",
      "20 20\n",
      "18 18\n",
      "13 13\n",
      "10 10\n",
      "8\n",
      "35 35\n",
      "30 30\n",
      "24 24\n",
      "20 20\n",
      "18 18\n",
      "13 13\n",
      "9\n",
      "33 33\n",
      "35 35\n",
      "30 30\n",
      "24 24\n",
      "20 20\n",
      "18 18\n",
      "10\n",
      "7 7\n",
      "11\n",
      "10 10\n",
      "7 7\n",
      "12\n",
      "11 11\n",
      "10 10\n",
      "7 7\n",
      "13\n",
      "12 12\n",
      "11 11\n",
      "10 10\n",
      "7 7\n",
      "14\n",
      "16 16\n",
      "12 12\n",
      "11 11\n",
      "10 10\n",
      "7 7\n",
      "15\n",
      "15 15\n",
      "16 16\n",
      "12 12\n",
      "11 11\n",
      "10 10\n",
      "7 7\n",
      "16\n",
      "12 12\n",
      "15 15\n",
      "16 16\n",
      "12 12\n",
      "11 11\n",
      "10 10\n",
      "17\n",
      "12 12\n",
      "12 12\n",
      "15 15\n",
      "16 16\n",
      "12 12\n",
      "11 11\n",
      "18\n",
      "15 15\n",
      "12 12\n",
      "12 12\n",
      "15 15\n",
      "16 16\n",
      "12 12\n",
      "19\n",
      "14 14\n",
      "15 15\n",
      "12 12\n",
      "12 12\n",
      "15 15\n",
      "16 16\n",
      "20\n",
      "15 15\n",
      "21\n",
      "8 8\n",
      "15 15\n",
      "22\n",
      "25 25\n",
      "8 8\n",
      "15 15\n",
      "23\n",
      "12 12\n",
      "25 25\n",
      "8 8\n",
      "15 15\n",
      "24\n",
      "24 24\n",
      "12 12\n",
      "25 25\n",
      "8 8\n",
      "15 15\n",
      "25\n",
      "19 19\n",
      "24 24\n",
      "12 12\n",
      "25 25\n",
      "8 8\n",
      "15 15\n",
      "26\n",
      "26 26\n",
      "19 19\n",
      "24 24\n",
      "12 12\n",
      "25 25\n",
      "8 8\n",
      "27\n",
      "29 29\n",
      "26 26\n",
      "19 19\n",
      "24 24\n",
      "12 12\n",
      "25 25\n",
      "28\n",
      "41 41\n",
      "29 29\n",
      "26 26\n",
      "19 19\n",
      "24 24\n",
      "12 12\n",
      "29\n",
      "59 59\n",
      "41 41\n",
      "29 29\n",
      "26 26\n",
      "19 19\n",
      "24 24\n",
      "30\n",
      "65 65\n",
      "59 59\n",
      "41 41\n",
      "29 29\n",
      "26 26\n",
      "19 19\n",
      "31\n",
      "58 58\n",
      "65 65\n",
      "59 59\n",
      "41 41\n",
      "29 29\n",
      "26 26\n",
      "32\n",
      "69 69\n",
      "58 58\n",
      "65 65\n",
      "59 59\n",
      "41 41\n",
      "29 29\n",
      "33\n",
      "94 94\n",
      "69 69\n",
      "58 58\n",
      "65 65\n",
      "59 59\n",
      "41 41\n",
      "34\n",
      "115 115\n",
      "94 94\n",
      "69 69\n",
      "58 58\n",
      "65 65\n",
      "59 59\n",
      "35\n",
      "121 121\n",
      "115 115\n",
      "94 94\n",
      "69 69\n",
      "58 58\n",
      "65 65\n",
      "36\n",
      "126 126\n",
      "121 121\n",
      "115 115\n",
      "94 94\n",
      "69 69\n",
      "58 58\n",
      "37\n",
      "138 138\n",
      "126 126\n",
      "121 121\n",
      "115 115\n",
      "94 94\n",
      "69 69\n",
      "38\n",
      "136 136\n",
      "138 138\n",
      "126 126\n",
      "121 121\n",
      "115 115\n",
      "94 94\n",
      "39\n",
      "145 145\n",
      "136 136\n",
      "138 138\n",
      "126 126\n",
      "121 121\n",
      "115 115\n",
      "40\n",
      "177 177\n",
      "145 145\n",
      "136 136\n",
      "138 138\n",
      "126 126\n",
      "121 121\n",
      "41\n",
      "173 173\n",
      "177 177\n",
      "145 145\n",
      "136 136\n",
      "138 138\n",
      "126 126\n",
      "42\n",
      "208 208\n",
      "173 173\n",
      "177 177\n",
      "145 145\n",
      "136 136\n",
      "138 138\n",
      "43\n",
      "224 224\n",
      "208 208\n",
      "173 173\n",
      "177 177\n",
      "145 145\n",
      "136 136\n",
      "44\n",
      "245 245\n",
      "224 224\n",
      "208 208\n",
      "173 173\n",
      "177 177\n",
      "145 145\n",
      "45\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "CAM_FRONT_RIGHT\n",
      "7 7\n",
      "CAM_FRONT_RIGHT\n",
      "7 7\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "CAM_FRONT_RIGHT\n",
      "7 7\n",
      "46\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "CAM_FRONT_RIGHT\n",
      "7 7\n",
      "CAM_FRONT_RIGHT\n",
      "7 7\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "47\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "CAM_FRONT_RIGHT\n",
      "7 7\n",
      "CAM_FRONT_RIGHT\n",
      "7 7\n",
      "48\n",
      "CAM_FRONT_RIGHT\n",
      "7 7\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "CAM_FRONT_RIGHT\n",
      "7 7\n",
      "49\n",
      "CAM_FRONT_RIGHT\n",
      "7 7\n",
      "CAM_FRONT_RIGHT\n",
      "7 7\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "50\n",
      "CAM_FRONT_RIGHT\n",
      "4 4\n",
      "CAM_FRONT_RIGHT\n",
      "7 7\n",
      "CAM_FRONT_RIGHT\n",
      "7 7\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "51\n",
      "CAM_FRONT_RIGHT\n",
      "3 3\n",
      "CAM_FRONT_RIGHT\n",
      "4 4\n",
      "CAM_FRONT_RIGHT\n",
      "7 7\n",
      "CAM_FRONT_RIGHT\n",
      "7 7\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "52\n",
      "CAM_FRONT_RIGHT\n",
      "0 0\n",
      "CAM_FRONT_RIGHT\n",
      "3 3\n",
      "CAM_FRONT_RIGHT\n",
      "4 4\n",
      "CAM_FRONT_RIGHT\n",
      "7 7\n",
      "CAM_FRONT_RIGHT\n",
      "7 7\n",
      "CAM_FRONT_RIGHT\n",
      "8 8\n",
      "53\n",
      "CAM_FRONT_RIGHT\n",
      "5 5\n",
      "CAM_FRONT_RIGHT\n",
      "0 0\n",
      "CAM_FRONT_RIGHT\n",
      "3 3\n",
      "CAM_FRONT_RIGHT\n",
      "4 4\n",
      "CAM_FRONT_RIGHT\n",
      "7 7\n",
      "CAM_FRONT_RIGHT\n",
      "7 7\n",
      "54\n",
      "CAM_FRONT_LEFT\n",
      "2 2\n",
      "55\n",
      "CAM_FRONT_LEFT\n",
      "4 4\n",
      "CAM_FRONT_LEFT\n",
      "2 2\n",
      "56\n",
      "CAM_FRONT_LEFT\n",
      "0 0\n",
      "CAM_FRONT_LEFT\n",
      "4 4\n",
      "CAM_FRONT_LEFT\n",
      "2 2\n",
      "57\n",
      "CAM_FRONT_LEFT\n",
      "2 2\n",
      "58\n",
      "CAM_FRONT_LEFT\n",
      "2 2\n",
      "CAM_FRONT_LEFT\n",
      "2 2\n",
      "59\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "CAM_FRONT_LEFT\n",
      "2 2\n",
      "CAM_FRONT_LEFT\n",
      "2 2\n",
      "60\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "CAM_FRONT_LEFT\n",
      "2 2\n",
      "CAM_FRONT_LEFT\n",
      "2 2\n",
      "61\n",
      "CAM_FRONT_LEFT\n",
      "4 4\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "CAM_FRONT_LEFT\n",
      "2 2\n",
      "CAM_FRONT_LEFT\n",
      "2 2\n",
      "62\n",
      "11 11\n",
      "CAM_FRONT_LEFT\n",
      "4 4\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "CAM_FRONT_LEFT\n",
      "2 2\n",
      "CAM_FRONT_LEFT\n",
      "2 2\n",
      "63\n",
      "6 6\n",
      "11 11\n",
      "CAM_FRONT_LEFT\n",
      "4 4\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "CAM_FRONT_LEFT\n",
      "2 2\n",
      "64\n",
      "15 15\n",
      "6 6\n",
      "11 11\n",
      "CAM_FRONT_LEFT\n",
      "4 4\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "65\n",
      "33 33\n",
      "15 15\n",
      "6 6\n",
      "11 11\n",
      "CAM_FRONT_LEFT\n",
      "4 4\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "66\n",
      "CAM_FRONT_RIGHT\n",
      "5 5\n",
      "CAM_FRONT_RIGHT\n",
      "5 5\n",
      "CAM_FRONT_RIGHT\n",
      "5 5\n",
      "CAM_FRONT_RIGHT\n",
      "5 5\n",
      "CAM_FRONT_RIGHT\n",
      "5 5\n",
      "CAM_FRONT_RIGHT\n",
      "5 5\n",
      "67\n",
      "CAM_FRONT_LEFT\n",
      "8 8\n",
      "68\n",
      "CAM_FRONT_LEFT\n",
      "11 11\n",
      "CAM_FRONT_LEFT\n",
      "8 8\n",
      "69\n",
      "CAM_FRONT_LEFT\n",
      "14 14\n",
      "CAM_FRONT_LEFT\n",
      "11 11\n",
      "CAM_FRONT_LEFT\n",
      "8 8\n",
      "70\n",
      "CAM_FRONT_LEFT\n",
      "15 15\n",
      "CAM_FRONT_LEFT\n",
      "14 14\n",
      "CAM_FRONT_LEFT\n",
      "11 11\n",
      "CAM_FRONT_LEFT\n",
      "8 8\n",
      "71\n",
      "CAM_FRONT_LEFT\n",
      "12 12\n",
      "CAM_FRONT_LEFT\n",
      "15 15\n",
      "CAM_FRONT_LEFT\n",
      "14 14\n",
      "CAM_FRONT_LEFT\n",
      "11 11\n",
      "CAM_FRONT_LEFT\n",
      "8 8\n",
      "72\n",
      "CAM_FRONT_LEFT\n",
      "14 14\n",
      "CAM_FRONT_LEFT\n",
      "12 12\n",
      "CAM_FRONT_LEFT\n",
      "15 15\n",
      "CAM_FRONT_LEFT\n",
      "14 14\n",
      "CAM_FRONT_LEFT\n",
      "11 11\n",
      "CAM_FRONT_LEFT\n",
      "8 8\n",
      "73\n",
      "CAM_FRONT_LEFT\n",
      "5 5\n",
      "CAM_FRONT_LEFT\n",
      "14 14\n",
      "CAM_FRONT_LEFT\n",
      "12 12\n",
      "CAM_FRONT_LEFT\n",
      "15 15\n",
      "CAM_FRONT_LEFT\n",
      "14 14\n",
      "CAM_FRONT_LEFT\n",
      "11 11\n",
      "74\n",
      "CAM_FRONT_LEFT\n",
      "0 0\n",
      "CAM_FRONT_LEFT\n",
      "5 5\n",
      "CAM_FRONT_LEFT\n",
      "14 14\n",
      "CAM_FRONT_LEFT\n",
      "12 12\n",
      "CAM_FRONT_LEFT\n",
      "15 15\n",
      "CAM_FRONT_LEFT\n",
      "14 14\n",
      "75\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "CAM_FRONT_LEFT\n",
      "0 0\n",
      "CAM_FRONT_LEFT\n",
      "5 5\n",
      "CAM_FRONT_LEFT\n",
      "14 14\n",
      "CAM_FRONT_LEFT\n",
      "12 12\n",
      "CAM_FRONT_LEFT\n",
      "15 15\n",
      "76\n",
      "CAM_FRONT_LEFT\n",
      "13 13\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "CAM_FRONT_LEFT\n",
      "0 0\n",
      "CAM_FRONT_LEFT\n",
      "5 5\n",
      "CAM_FRONT_LEFT\n",
      "14 14\n",
      "CAM_FRONT_LEFT\n",
      "12 12\n",
      "77\n",
      "61 61\n",
      "CAM_FRONT_LEFT\n",
      "13 13\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "CAM_FRONT_LEFT\n",
      "0 0\n",
      "CAM_FRONT_LEFT\n",
      "5 5\n",
      "CAM_FRONT_LEFT\n",
      "14 14\n",
      "78\n",
      "72 72\n",
      "61 61\n",
      "CAM_FRONT_LEFT\n",
      "13 13\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "CAM_FRONT_LEFT\n",
      "0 0\n",
      "CAM_FRONT_LEFT\n",
      "5 5\n",
      "79\n",
      "64 64\n",
      "72 72\n",
      "61 61\n",
      "CAM_FRONT_LEFT\n",
      "13 13\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "CAM_FRONT_LEFT\n",
      "0 0\n",
      "80\n",
      "73 73\n",
      "64 64\n",
      "72 72\n",
      "61 61\n",
      "CAM_FRONT_LEFT\n",
      "13 13\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "81\n",
      "120 120\n",
      "73 73\n",
      "64 64\n",
      "72 72\n",
      "61 61\n",
      "CAM_FRONT_LEFT\n",
      "13 13\n",
      "82\n",
      "151 151\n",
      "120 120\n",
      "73 73\n",
      "64 64\n",
      "72 72\n",
      "61 61\n",
      "83\n",
      "126 126\n",
      "151 151\n",
      "120 120\n",
      "73 73\n",
      "64 64\n",
      "72 72\n",
      "84\n",
      "78 78\n",
      "126 126\n",
      "151 151\n",
      "120 120\n",
      "73 73\n",
      "64 64\n",
      "85\n",
      "61 61\n",
      "78 78\n",
      "126 126\n",
      "151 151\n",
      "120 120\n",
      "73 73\n",
      "86\n",
      "31 31\n",
      "61 61\n",
      "78 78\n",
      "126 126\n",
      "151 151\n",
      "120 120\n",
      "87\n",
      "18 18\n",
      "31 31\n",
      "61 61\n",
      "78 78\n",
      "126 126\n",
      "151 151\n",
      "88\n",
      "14 14\n",
      "18 18\n",
      "31 31\n",
      "61 61\n",
      "78 78\n",
      "126 126\n",
      "89\n",
      "14 14\n",
      "14 14\n",
      "18 18\n",
      "31 31\n",
      "61 61\n",
      "78 78\n",
      "90\n",
      "12 12\n",
      "14 14\n",
      "14 14\n",
      "18 18\n",
      "31 31\n",
      "61 61\n",
      "91\n",
      "13 13\n",
      "12 12\n",
      "14 14\n",
      "14 14\n",
      "18 18\n",
      "31 31\n",
      "92\n",
      "15 15\n",
      "13 13\n",
      "12 12\n",
      "14 14\n",
      "14 14\n",
      "18 18\n",
      "93\n",
      "CAM_FRONT_LEFT\n",
      "0 0\n",
      "94\n",
      "CAM_FRONT_LEFT\n",
      "5 5\n",
      "CAM_FRONT_LEFT\n",
      "0 0\n",
      "95\n",
      "CAM_FRONT_LEFT\n",
      "12 12\n",
      "CAM_FRONT_LEFT\n",
      "5 5\n",
      "CAM_FRONT_LEFT\n",
      "0 0\n",
      "96\n",
      "CAM_FRONT_LEFT\n",
      "9 9\n",
      "CAM_FRONT_LEFT\n",
      "12 12\n",
      "CAM_FRONT_LEFT\n",
      "5 5\n",
      "CAM_FRONT_LEFT\n",
      "0 0\n",
      "97\n",
      "CAM_FRONT_LEFT\n",
      "27 27\n",
      "CAM_FRONT_LEFT\n",
      "9 9\n",
      "CAM_FRONT_LEFT\n",
      "12 12\n",
      "CAM_FRONT_LEFT\n",
      "5 5\n",
      "CAM_FRONT_LEFT\n",
      "0 0\n",
      "98\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "99\n",
      "CAM_FRONT_LEFT\n",
      "0 0\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "100\n",
      "CAM_FRONT_LEFT\n",
      "8 8\n",
      "CAM_FRONT_LEFT\n",
      "0 0\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "101\n",
      "CAM_FRONT_LEFT\n",
      "11 11\n",
      "CAM_FRONT_LEFT\n",
      "8 8\n",
      "CAM_FRONT_LEFT\n",
      "0 0\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "102\n",
      "CAM_FRONT_LEFT\n",
      "5 5\n",
      "CAM_FRONT_LEFT\n",
      "11 11\n",
      "CAM_FRONT_LEFT\n",
      "8 8\n",
      "CAM_FRONT_LEFT\n",
      "0 0\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "103\n",
      "CAM_FRONT_LEFT\n",
      "27 27\n",
      "CAM_FRONT_LEFT\n",
      "5 5\n",
      "CAM_FRONT_LEFT\n",
      "11 11\n",
      "CAM_FRONT_LEFT\n",
      "8 8\n",
      "CAM_FRONT_LEFT\n",
      "0 0\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "104\n",
      "72 72\n",
      "CAM_FRONT_LEFT\n",
      "27 27\n",
      "CAM_FRONT_LEFT\n",
      "5 5\n",
      "CAM_FRONT_LEFT\n",
      "11 11\n",
      "CAM_FRONT_LEFT\n",
      "8 8\n",
      "CAM_FRONT_LEFT\n",
      "0 0\n",
      "105\n",
      "124 124\n",
      "72 72\n",
      "CAM_FRONT_LEFT\n",
      "27 27\n",
      "CAM_FRONT_LEFT\n",
      "5 5\n",
      "CAM_FRONT_LEFT\n",
      "11 11\n",
      "CAM_FRONT_LEFT\n",
      "8 8\n",
      "106\n",
      "148 148\n",
      "124 124\n",
      "72 72\n",
      "CAM_FRONT_LEFT\n",
      "27 27\n",
      "CAM_FRONT_LEFT\n",
      "5 5\n",
      "CAM_FRONT_LEFT\n",
      "11 11\n",
      "107\n",
      "125 125\n",
      "148 148\n",
      "124 124\n",
      "72 72\n",
      "CAM_FRONT_LEFT\n",
      "27 27\n",
      "CAM_FRONT_LEFT\n",
      "5 5\n",
      "108\n",
      "140 140\n",
      "125 125\n",
      "148 148\n",
      "124 124\n",
      "72 72\n",
      "CAM_FRONT_LEFT\n",
      "27 27\n",
      "109\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "110\n",
      "CAM_FRONT_LEFT\n",
      "7 7\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "111\n",
      "CAM_FRONT_LEFT\n",
      "10 10\n",
      "CAM_FRONT_LEFT\n",
      "7 7\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "112\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "CAM_FRONT_LEFT\n",
      "10 10\n",
      "CAM_FRONT_LEFT\n",
      "7 7\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "113\n",
      "CAM_FRONT_LEFT\n",
      "7 7\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "CAM_FRONT_LEFT\n",
      "10 10\n",
      "CAM_FRONT_LEFT\n",
      "7 7\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "114\n",
      "12 12\n",
      "CAM_FRONT_LEFT\n",
      "7 7\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "CAM_FRONT_LEFT\n",
      "10 10\n",
      "CAM_FRONT_LEFT\n",
      "7 7\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "115\n",
      "32 32\n",
      "12 12\n",
      "CAM_FRONT_LEFT\n",
      "7 7\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "CAM_FRONT_LEFT\n",
      "10 10\n",
      "CAM_FRONT_LEFT\n",
      "7 7\n",
      "116\n",
      "67 67\n",
      "32 32\n",
      "12 12\n",
      "CAM_FRONT_LEFT\n",
      "7 7\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "CAM_FRONT_LEFT\n",
      "10 10\n",
      "117\n",
      "64 64\n",
      "67 67\n",
      "32 32\n",
      "12 12\n",
      "CAM_FRONT_LEFT\n",
      "7 7\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "118\n",
      "28 28\n",
      "64 64\n",
      "67 67\n",
      "32 32\n",
      "12 12\n",
      "CAM_FRONT_LEFT\n",
      "7 7\n",
      "119\n",
      "34 34\n",
      "28 28\n",
      "64 64\n",
      "67 67\n",
      "32 32\n",
      "12 12\n",
      "120\n",
      "58 58\n",
      "34 34\n",
      "28 28\n",
      "64 64\n",
      "67 67\n",
      "32 32\n",
      "121\n",
      "CAM_FRONT_RIGHT\n",
      "35 35\n",
      "58 58\n",
      "34 34\n",
      "28 28\n",
      "64 64\n",
      "67 67\n",
      "122\n",
      "0 0\n",
      "123\n",
      "0 0\n",
      "0 0\n",
      "124\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "125\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "126\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "127\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "128\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "129\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "130\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "131\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "132\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "133\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "134\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "135\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "136\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "137\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "138\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "139\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "140\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "141\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "142\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "143\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "144\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "145\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "146\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "147\n",
      "CAM_FRONT_LEFT\n",
      "7 7\n",
      "148\n",
      "CAM_FRONT_LEFT\n",
      "5 5\n",
      "CAM_FRONT_LEFT\n",
      "7 7\n",
      "149\n",
      "CAM_FRONT_LEFT\n",
      "4 4\n",
      "150\n",
      "CAM_FRONT_LEFT\n",
      "4 4\n",
      "CAM_FRONT_LEFT\n",
      "4 4\n",
      "151\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "CAM_FRONT_LEFT\n",
      "4 4\n",
      "CAM_FRONT_LEFT\n",
      "4 4\n",
      "152\n",
      "CAM_FRONT_LEFT\n",
      "8 8\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "CAM_FRONT_LEFT\n",
      "4 4\n",
      "CAM_FRONT_LEFT\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4 4\n",
      "153\n",
      "CAM_FRONT_LEFT\n",
      "18 18\n",
      "CAM_FRONT_LEFT\n",
      "8 8\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "CAM_FRONT_LEFT\n",
      "4 4\n",
      "CAM_FRONT_LEFT\n",
      "4 4\n",
      "154\n",
      "CAM_FRONT_LEFT\n",
      "22 22\n",
      "CAM_FRONT_LEFT\n",
      "18 18\n",
      "CAM_FRONT_LEFT\n",
      "8 8\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "CAM_FRONT_LEFT\n",
      "4 4\n",
      "CAM_FRONT_LEFT\n",
      "4 4\n",
      "155\n",
      "CAM_FRONT_LEFT\n",
      "40 40\n",
      "CAM_FRONT_LEFT\n",
      "22 22\n",
      "CAM_FRONT_LEFT\n",
      "18 18\n",
      "CAM_FRONT_LEFT\n",
      "8 8\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "CAM_FRONT_LEFT\n",
      "4 4\n",
      "156\n",
      "CAM_FRONT_LEFT\n",
      "43 43\n",
      "CAM_FRONT_LEFT\n",
      "40 40\n",
      "CAM_FRONT_LEFT\n",
      "22 22\n",
      "CAM_FRONT_LEFT\n",
      "18 18\n",
      "CAM_FRONT_LEFT\n",
      "8 8\n",
      "CAM_FRONT_LEFT\n",
      "6 6\n",
      "157\n",
      "CAM_FRONT_LEFT\n",
      "97 97\n",
      "CAM_FRONT_LEFT\n",
      "43 43\n",
      "CAM_FRONT_LEFT\n",
      "40 40\n",
      "CAM_FRONT_LEFT\n",
      "22 22\n",
      "CAM_FRONT_LEFT\n",
      "18 18\n",
      "CAM_FRONT_LEFT\n",
      "8 8\n",
      "158\n",
      "156 156\n",
      "CAM_FRONT_LEFT\n",
      "97 97\n",
      "CAM_FRONT_LEFT\n",
      "43 43\n",
      "CAM_FRONT_LEFT\n",
      "40 40\n",
      "CAM_FRONT_LEFT\n",
      "22 22\n",
      "CAM_FRONT_LEFT\n",
      "18 18\n",
      "159\n",
      "229 229\n",
      "156 156\n",
      "CAM_FRONT_LEFT\n",
      "97 97\n",
      "CAM_FRONT_LEFT\n",
      "43 43\n",
      "CAM_FRONT_LEFT\n",
      "40 40\n",
      "CAM_FRONT_LEFT\n",
      "22 22\n",
      "160\n",
      "284 284\n",
      "229 229\n",
      "156 156\n",
      "CAM_FRONT_LEFT\n",
      "97 97\n",
      "CAM_FRONT_LEFT\n",
      "43 43\n",
      "CAM_FRONT_LEFT\n",
      "40 40\n",
      "161\n",
      "0 0\n",
      "162\n",
      "0 0\n",
      "0 0\n",
      "163\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "164\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "165\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "166\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "167\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "168\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "169\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "170\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "171\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "172\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "173\n",
      "3 3\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "174\n",
      "1 1\n",
      "3 3\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "175\n",
      "2 2\n",
      "1 1\n",
      "3 3\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "176\n",
      "0 0\n",
      "2 2\n",
      "1 1\n",
      "3 3\n",
      "0 0\n",
      "0 0\n",
      "177\n",
      "0 0\n",
      "0 0\n",
      "2 2\n",
      "1 1\n",
      "3 3\n",
      "0 0\n",
      "178\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "2 2\n",
      "1 1\n",
      "3 3\n",
      "179\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "2 2\n",
      "1 1\n",
      "180\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "181\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "182\n",
      "CAM_FRONT_LEFT\n",
      "3 3\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "183\n",
      "CAM_FRONT_LEFT\n",
      "2 2\n",
      "CAM_FRONT_LEFT\n",
      "3 3\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "184\n",
      "19 19\n",
      "CAM_FRONT_LEFT\n",
      "2 2\n",
      "CAM_FRONT_LEFT\n",
      "3 3\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "185\n",
      "33 33\n",
      "19 19\n",
      "CAM_FRONT_LEFT\n",
      "2 2\n",
      "CAM_FRONT_LEFT\n",
      "3 3\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "186\n",
      "74 74\n",
      "33 33\n",
      "19 19\n",
      "CAM_FRONT_LEFT\n",
      "2 2\n",
      "CAM_FRONT_LEFT\n",
      "3 3\n",
      "CAM_FRONT_LEFT\n",
      "1 1\n",
      "187\n",
      "73 73\n",
      "74 74\n",
      "33 33\n",
      "19 19\n",
      "CAM_FRONT_LEFT\n",
      "2 2\n",
      "CAM_FRONT_LEFT\n",
      "3 3\n",
      "188\n",
      "46 46\n",
      "73 73\n",
      "74 74\n",
      "33 33\n",
      "19 19\n",
      "CAM_FRONT_LEFT\n",
      "2 2\n",
      "189\n",
      "28 28\n",
      "46 46\n",
      "73 73\n",
      "74 74\n",
      "33 33\n",
      "19 19\n",
      "190\n",
      "CAM_BACK\n",
      "6 6\n",
      "CAM_BACK\n",
      "7 7\n",
      "CAM_BACK\n",
      "6 6\n",
      "CAM_BACK\n",
      "10 10\n",
      "CAM_BACK\n",
      "6 6\n",
      "191\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "6 6\n",
      "CAM_BACK\n",
      "7 7\n",
      "CAM_BACK\n",
      "6 6\n",
      "CAM_BACK\n",
      "10 10\n",
      "CAM_BACK\n",
      "6 6\n",
      "192\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "6 6\n",
      "CAM_BACK\n",
      "7 7\n",
      "CAM_BACK\n",
      "6 6\n",
      "CAM_BACK\n",
      "10 10\n",
      "193\n",
      "CAM_BACK\n",
      "6 6\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "6 6\n",
      "CAM_BACK\n",
      "7 7\n",
      "CAM_BACK\n",
      "6 6\n",
      "194\n",
      "CAM_BACK\n",
      "6 6\n",
      "CAM_BACK\n",
      "6 6\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "6 6\n",
      "CAM_BACK\n",
      "7 7\n",
      "195\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "6 6\n",
      "CAM_BACK\n",
      "6 6\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "6 6\n",
      "196\n",
      "CAM_BACK\n",
      "2 2\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "6 6\n",
      "CAM_BACK\n",
      "6 6\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "4 4\n",
      "197\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "2 2\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "6 6\n",
      "CAM_BACK\n",
      "6 6\n",
      "CAM_BACK\n",
      "3 3\n",
      "198\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "2 2\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "6 6\n",
      "CAM_BACK\n",
      "6 6\n",
      "199\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "2 2\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "6 6\n",
      "200\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "2 2\n",
      "CAM_BACK\n",
      "3 3\n",
      "201\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "2 2\n",
      "202\n",
      "CAM_BACK\n",
      "2 2\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "4 4\n",
      "203\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "2 2\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "0 0\n",
      "204\n",
      "CAM_BACK\n",
      "2 2\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "2 2\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "3 3\n",
      "205\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "2 2\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "2 2\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "1 1\n",
      "206\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "2 2\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "2 2\n",
      "CAM_BACK\n",
      "1 1\n",
      "207\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "2 2\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "2 2\n",
      "208\n",
      "CAM_BACK\n",
      "5 5\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "2 2\n",
      "CAM_BACK\n",
      "4 4\n",
      "209\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "5 5\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "2 2\n",
      "210\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "5 5\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "3 3\n",
      "211\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "5 5\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "0 0\n",
      "212\n",
      "CAM_BACK\n",
      "5 5\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "5 5\n",
      "CAM_BACK\n",
      "3 3\n",
      "213\n",
      "CAM_BACK\n",
      "20 20\n",
      "214\n",
      "CAM_BACK\n",
      "32 32\n",
      "CAM_BACK\n",
      "20 20\n",
      "215\n",
      "CAM_BACK\n",
      "15 15\n",
      "CAM_BACK\n",
      "32 32\n",
      "CAM_BACK\n",
      "20 20\n",
      "216\n",
      "CAM_BACK\n",
      "34 34\n",
      "CAM_BACK\n",
      "15 15\n",
      "CAM_BACK\n",
      "32 32\n",
      "CAM_BACK\n",
      "20 20\n",
      "217\n",
      "CAM_BACK\n",
      "15 15\n",
      "CAM_BACK\n",
      "34 34\n",
      "CAM_BACK\n",
      "15 15\n",
      "CAM_BACK\n",
      "32 32\n",
      "CAM_BACK\n",
      "20 20\n",
      "218\n",
      "CAM_BACK\n",
      "15 15\n",
      "CAM_BACK\n",
      "15 15\n",
      "CAM_BACK\n",
      "34 34\n",
      "CAM_BACK\n",
      "15 15\n",
      "CAM_BACK\n",
      "32 32\n",
      "CAM_BACK\n",
      "20 20\n",
      "219\n",
      "CAM_BACK\n",
      "13 13\n",
      "CAM_BACK\n",
      "15 15\n",
      "CAM_BACK\n",
      "15 15\n",
      "CAM_BACK\n",
      "34 34\n",
      "CAM_BACK\n",
      "15 15\n",
      "CAM_BACK\n",
      "32 32\n",
      "220\n",
      "CAM_BACK\n",
      "13 13\n",
      "CAM_BACK\n",
      "13 13\n",
      "CAM_BACK\n",
      "15 15\n",
      "CAM_BACK\n",
      "15 15\n",
      "CAM_BACK\n",
      "34 34\n",
      "CAM_BACK\n",
      "15 15\n",
      "221\n",
      "CAM_BACK\n",
      "11 11\n",
      "CAM_BACK\n",
      "13 13\n",
      "CAM_BACK\n",
      "13 13\n",
      "CAM_BACK\n",
      "15 15\n",
      "CAM_BACK\n",
      "15 15\n",
      "CAM_BACK\n",
      "34 34\n",
      "222\n",
      "CAM_BACK\n",
      "17 17\n",
      "CAM_BACK\n",
      "11 11\n",
      "CAM_BACK\n",
      "13 13\n",
      "CAM_BACK\n",
      "13 13\n",
      "CAM_BACK\n",
      "15 15\n",
      "CAM_BACK\n",
      "15 15\n",
      "223\n",
      "CAM_BACK\n",
      "35 35\n",
      "CAM_BACK\n",
      "17 17\n",
      "CAM_BACK\n",
      "11 11\n",
      "CAM_BACK\n",
      "13 13\n",
      "CAM_BACK\n",
      "13 13\n",
      "CAM_BACK\n",
      "15 15\n",
      "224\n",
      "CAM_BACK\n",
      "52 52\n",
      "CAM_BACK\n",
      "35 35\n",
      "CAM_BACK\n",
      "17 17\n",
      "CAM_BACK\n",
      "11 11\n",
      "CAM_BACK\n",
      "13 13\n",
      "CAM_BACK\n",
      "13 13\n",
      "225\n",
      "CAM_BACK\n",
      "68 68\n",
      "CAM_BACK\n",
      "52 52\n",
      "CAM_BACK\n",
      "35 35\n",
      "CAM_BACK\n",
      "17 17\n",
      "CAM_BACK\n",
      "11 11\n",
      "CAM_BACK\n",
      "13 13\n",
      "226\n",
      "CAM_BACK\n",
      "85 85\n",
      "CAM_BACK\n",
      "68 68\n",
      "CAM_BACK\n",
      "52 52\n",
      "CAM_BACK\n",
      "35 35\n",
      "CAM_BACK\n",
      "17 17\n",
      "CAM_BACK\n",
      "11 11\n",
      "227\n",
      "CAM_BACK\n",
      "103 103\n",
      "CAM_BACK\n",
      "85 85\n",
      "CAM_BACK\n",
      "68 68\n",
      "CAM_BACK\n",
      "52 52\n",
      "CAM_BACK\n",
      "35 35\n",
      "CAM_BACK\n",
      "17 17\n",
      "228\n",
      "CAM_BACK\n",
      "110 110\n",
      "CAM_BACK\n",
      "103 103\n",
      "CAM_BACK\n",
      "85 85\n",
      "CAM_BACK\n",
      "68 68\n",
      "CAM_BACK\n",
      "52 52\n",
      "CAM_BACK\n",
      "35 35\n",
      "229\n",
      "CAM_BACK\n",
      "119 119\n",
      "CAM_BACK\n",
      "110 110\n",
      "CAM_BACK\n",
      "103 103\n",
      "CAM_BACK\n",
      "85 85\n",
      "CAM_BACK\n",
      "68 68\n",
      "CAM_BACK\n",
      "52 52\n",
      "230\n",
      "CAM_BACK\n",
      "188 188\n",
      "CAM_BACK\n",
      "119 119\n",
      "CAM_BACK\n",
      "110 110\n",
      "CAM_BACK\n",
      "103 103\n",
      "CAM_BACK\n",
      "85 85\n",
      "CAM_BACK\n",
      "68 68\n",
      "231\n",
      "CAM_BACK\n",
      "278 278\n",
      "CAM_BACK\n",
      "188 188\n",
      "CAM_BACK\n",
      "119 119\n",
      "CAM_BACK\n",
      "110 110\n",
      "CAM_BACK\n",
      "103 103\n",
      "CAM_BACK\n",
      "85 85\n",
      "232\n",
      "CAM_BACK\n",
      "468 468\n",
      "CAM_BACK\n",
      "278 278\n",
      "CAM_BACK\n",
      "188 188\n",
      "CAM_BACK\n",
      "119 119\n",
      "CAM_BACK\n",
      "110 110\n",
      "CAM_BACK\n",
      "103 103\n",
      "233\n",
      "CAM_BACK\n",
      "912 912\n",
      "CAM_BACK\n",
      "468 468\n",
      "CAM_BACK\n",
      "278 278\n",
      "CAM_BACK\n",
      "188 188\n",
      "CAM_BACK\n",
      "119 119\n",
      "CAM_BACK\n",
      "110 110\n",
      "234\n",
      "CAM_BACK\n",
      "1878 1878\n",
      "CAM_BACK\n",
      "912 912\n",
      "CAM_BACK\n",
      "468 468\n",
      "CAM_BACK\n",
      "278 278\n",
      "CAM_BACK\n",
      "188 188\n",
      "CAM_BACK\n",
      "119 119\n",
      "235\n",
      "CAM_BACK\n",
      "3472 3472\n",
      "CAM_BACK\n",
      "1878 1878\n",
      "CAM_BACK\n",
      "912 912\n",
      "CAM_BACK\n",
      "468 468\n",
      "CAM_BACK\n",
      "278 278\n",
      "CAM_BACK\n",
      "188 188\n",
      "236\n",
      "CAM_FRONT_RIGHT\n",
      "2 2\n",
      "237\n",
      "CAM_BACK\n",
      "134 134\n",
      "238\n",
      "CAM_BACK\n",
      "170 170\n",
      "CAM_BACK\n",
      "134 134\n",
      "239\n",
      "CAM_BACK\n",
      "221 221\n",
      "CAM_BACK\n",
      "170 170\n",
      "CAM_BACK\n",
      "134 134\n",
      "240\n",
      "CAM_BACK\n",
      "215 215\n",
      "CAM_BACK\n",
      "221 221\n",
      "CAM_BACK\n",
      "170 170\n",
      "CAM_BACK\n",
      "134 134\n",
      "241\n",
      "CAM_BACK\n",
      "226 226\n",
      "CAM_BACK\n",
      "215 215\n",
      "CAM_BACK\n",
      "221 221\n",
      "CAM_BACK\n",
      "170 170\n",
      "CAM_BACK\n",
      "134 134\n",
      "242\n",
      "CAM_BACK\n",
      "220 220\n",
      "CAM_BACK\n",
      "226 226\n",
      "CAM_BACK\n",
      "215 215\n",
      "CAM_BACK\n",
      "221 221\n",
      "CAM_BACK\n",
      "170 170\n",
      "CAM_BACK\n",
      "134 134\n",
      "243\n",
      "CAM_BACK\n",
      "230 230\n",
      "CAM_BACK\n",
      "220 220\n",
      "CAM_BACK\n",
      "226 226\n",
      "CAM_BACK\n",
      "215 215\n",
      "CAM_BACK\n",
      "221 221\n",
      "CAM_BACK\n",
      "170 170\n",
      "244\n",
      "CAM_BACK\n",
      "241 241\n",
      "CAM_BACK\n",
      "230 230\n",
      "CAM_BACK\n",
      "220 220\n",
      "CAM_BACK\n",
      "226 226\n",
      "CAM_BACK\n",
      "215 215\n",
      "CAM_BACK\n",
      "221 221\n",
      "245\n",
      "CAM_BACK\n",
      "275 275\n",
      "CAM_BACK\n",
      "241 241\n",
      "CAM_BACK\n",
      "230 230\n",
      "CAM_BACK\n",
      "220 220\n",
      "CAM_BACK\n",
      "226 226\n",
      "CAM_BACK\n",
      "215 215\n",
      "246\n",
      "CAM_BACK\n",
      "339 339\n",
      "CAM_BACK\n",
      "275 275\n",
      "CAM_BACK\n",
      "241 241\n",
      "CAM_BACK\n",
      "230 230\n",
      "CAM_BACK\n",
      "220 220\n",
      "CAM_BACK\n",
      "226 226\n",
      "247\n",
      "CAM_BACK\n",
      "320 320\n",
      "CAM_BACK\n",
      "339 339\n",
      "CAM_BACK\n",
      "275 275\n",
      "CAM_BACK\n",
      "241 241\n",
      "CAM_BACK\n",
      "230 230\n",
      "CAM_BACK\n",
      "220 220\n",
      "248\n",
      "CAM_BACK\n",
      "374 374\n",
      "CAM_BACK\n",
      "320 320\n",
      "CAM_BACK\n",
      "339 339\n",
      "CAM_BACK\n",
      "275 275\n",
      "CAM_BACK\n",
      "241 241\n",
      "CAM_BACK\n",
      "230 230\n",
      "249\n",
      "CAM_BACK\n",
      "405 405\n",
      "CAM_BACK\n",
      "374 374\n",
      "CAM_BACK\n",
      "320 320\n",
      "CAM_BACK\n",
      "339 339\n",
      "CAM_BACK\n",
      "275 275\n",
      "CAM_BACK\n",
      "241 241\n",
      "250\n",
      "CAM_BACK\n",
      "427 427\n",
      "CAM_BACK\n",
      "405 405\n",
      "CAM_BACK\n",
      "374 374\n",
      "CAM_BACK\n",
      "320 320\n",
      "CAM_BACK\n",
      "339 339\n",
      "CAM_BACK\n",
      "275 275\n",
      "251\n",
      "3 3\n",
      "252\n",
      "13 13\n",
      "3 3\n",
      "253\n",
      "7 7\n",
      "13 13\n",
      "3 3\n",
      "254\n",
      "25 25\n",
      "7 7\n",
      "13 13\n",
      "3 3\n",
      "255\n",
      "74 74\n",
      "25 25\n",
      "7 7\n",
      "13 13\n",
      "3 3\n",
      "256\n",
      "113 113\n",
      "74 74\n",
      "25 25\n",
      "7 7\n",
      "13 13\n",
      "3 3\n",
      "257\n",
      "171 171\n",
      "113 113\n",
      "74 74\n",
      "25 25\n",
      "7 7\n",
      "13 13\n",
      "258\n",
      "145 145\n",
      "171 171\n",
      "113 113\n",
      "74 74\n",
      "25 25\n",
      "7 7\n",
      "259\n",
      "223 223\n",
      "145 145\n",
      "171 171\n",
      "113 113\n",
      "74 74\n",
      "25 25\n",
      "260\n",
      "273 273\n",
      "223 223\n",
      "145 145\n",
      "171 171\n",
      "113 113\n",
      "74 74\n",
      "261\n",
      "264 264\n",
      "273 273\n",
      "223 223\n",
      "145 145\n",
      "171 171\n",
      "113 113\n",
      "262\n",
      "CAM_FRONT_RIGHT\n",
      "1 1\n",
      "263\n",
      "CAM_FRONT_RIGHT\n",
      "6 6\n",
      "CAM_FRONT_RIGHT\n",
      "1 1\n",
      "264\n",
      "CAM_BACK\n",
      "6 6\n",
      "265\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "6 6\n",
      "266\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "6 6\n",
      "267\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "6 6\n",
      "268\n",
      "CAM_BACK\n",
      "8 8\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "6 6\n",
      "269\n",
      "CAM_BACK\n",
      "9 9\n",
      "CAM_BACK\n",
      "8 8\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "6 6\n",
      "270\n",
      "CAM_BACK\n",
      "9 9\n",
      "CAM_BACK\n",
      "9 9\n",
      "CAM_BACK\n",
      "8 8\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "1 1\n",
      "271\n",
      "CAM_BACK\n",
      "8 8\n",
      "CAM_BACK\n",
      "9 9\n",
      "CAM_BACK\n",
      "9 9\n",
      "CAM_BACK\n",
      "8 8\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "0 0\n",
      "272\n",
      "CAM_BACK\n",
      "10 10\n",
      "CAM_BACK\n",
      "8 8\n",
      "CAM_BACK\n",
      "9 9\n",
      "CAM_BACK\n",
      "9 9\n",
      "CAM_BACK\n",
      "8 8\n",
      "CAM_BACK\n",
      "0 0\n",
      "273\n",
      "CAM_BACK\n",
      "8 8\n",
      "CAM_BACK\n",
      "10 10\n",
      "CAM_BACK\n",
      "8 8\n",
      "CAM_BACK\n",
      "9 9\n",
      "CAM_BACK\n",
      "9 9\n",
      "CAM_BACK\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "8 8\n",
      "274\n",
      "CAM_BACK\n",
      "20 20\n",
      "CAM_BACK\n",
      "8 8\n",
      "CAM_BACK\n",
      "10 10\n",
      "CAM_BACK\n",
      "8 8\n",
      "CAM_BACK\n",
      "9 9\n",
      "CAM_BACK\n",
      "9 9\n",
      "275\n",
      "CAM_BACK\n",
      "13 13\n",
      "CAM_BACK\n",
      "20 20\n",
      "CAM_BACK\n",
      "8 8\n",
      "CAM_BACK\n",
      "10 10\n",
      "CAM_BACK\n",
      "8 8\n",
      "CAM_BACK\n",
      "9 9\n",
      "276\n",
      "CAM_BACK\n",
      "13 13\n",
      "CAM_BACK\n",
      "13 13\n",
      "CAM_BACK\n",
      "20 20\n",
      "CAM_BACK\n",
      "8 8\n",
      "CAM_BACK\n",
      "10 10\n",
      "CAM_BACK\n",
      "8 8\n",
      "277\n",
      "CAM_BACK\n",
      "34 34\n",
      "CAM_BACK\n",
      "13 13\n",
      "CAM_BACK\n",
      "13 13\n",
      "CAM_BACK\n",
      "20 20\n",
      "CAM_BACK\n",
      "8 8\n",
      "CAM_BACK\n",
      "10 10\n",
      "278\n",
      "CAM_BACK\n",
      "23 23\n",
      "CAM_BACK\n",
      "34 34\n",
      "CAM_BACK\n",
      "13 13\n",
      "CAM_BACK\n",
      "13 13\n",
      "CAM_BACK\n",
      "20 20\n",
      "CAM_BACK\n",
      "8 8\n",
      "279\n",
      "CAM_BACK\n",
      "32 32\n",
      "CAM_BACK\n",
      "23 23\n",
      "CAM_BACK\n",
      "34 34\n",
      "CAM_BACK\n",
      "13 13\n",
      "CAM_BACK\n",
      "13 13\n",
      "CAM_BACK\n",
      "20 20\n",
      "280\n",
      "CAM_BACK\n",
      "39 39\n",
      "CAM_BACK\n",
      "32 32\n",
      "CAM_BACK\n",
      "23 23\n",
      "CAM_BACK\n",
      "34 34\n",
      "CAM_BACK\n",
      "13 13\n",
      "CAM_BACK\n",
      "13 13\n",
      "281\n",
      "CAM_BACK\n",
      "34 34\n",
      "CAM_BACK\n",
      "39 39\n",
      "CAM_BACK\n",
      "32 32\n",
      "CAM_BACK\n",
      "23 23\n",
      "CAM_BACK\n",
      "34 34\n",
      "CAM_BACK\n",
      "13 13\n",
      "282\n",
      "CAM_BACK\n",
      "59 59\n",
      "CAM_BACK\n",
      "34 34\n",
      "CAM_BACK\n",
      "39 39\n",
      "CAM_BACK\n",
      "32 32\n",
      "CAM_BACK\n",
      "23 23\n",
      "CAM_BACK\n",
      "34 34\n",
      "283\n",
      "CAM_BACK\n",
      "70 70\n",
      "CAM_BACK\n",
      "59 59\n",
      "CAM_BACK\n",
      "34 34\n",
      "CAM_BACK\n",
      "39 39\n",
      "CAM_BACK\n",
      "32 32\n",
      "CAM_BACK\n",
      "23 23\n",
      "284\n",
      "CAM_BACK\n",
      "59 59\n",
      "CAM_BACK\n",
      "70 70\n",
      "CAM_BACK\n",
      "59 59\n",
      "CAM_BACK\n",
      "34 34\n",
      "CAM_BACK\n",
      "39 39\n",
      "CAM_BACK\n",
      "32 32\n",
      "285\n",
      "CAM_BACK\n",
      "63 63\n",
      "CAM_BACK\n",
      "59 59\n",
      "CAM_BACK\n",
      "70 70\n",
      "CAM_BACK\n",
      "59 59\n",
      "CAM_BACK\n",
      "34 34\n",
      "CAM_BACK\n",
      "39 39\n",
      "286\n",
      "CAM_BACK\n",
      "84 84\n",
      "CAM_BACK\n",
      "63 63\n",
      "CAM_BACK\n",
      "59 59\n",
      "CAM_BACK\n",
      "70 70\n",
      "CAM_BACK\n",
      "59 59\n",
      "CAM_BACK\n",
      "34 34\n",
      "287\n",
      "CAM_BACK\n",
      "127 127\n",
      "CAM_BACK\n",
      "84 84\n",
      "CAM_BACK\n",
      "63 63\n",
      "CAM_BACK\n",
      "59 59\n",
      "CAM_BACK\n",
      "70 70\n",
      "CAM_BACK\n",
      "59 59\n",
      "288\n",
      "CAM_BACK\n",
      "144 144\n",
      "CAM_BACK\n",
      "127 127\n",
      "CAM_BACK\n",
      "84 84\n",
      "CAM_BACK\n",
      "63 63\n",
      "CAM_BACK\n",
      "59 59\n",
      "CAM_BACK\n",
      "70 70\n",
      "289\n",
      "CAM_BACK\n",
      "222 222\n",
      "CAM_BACK\n",
      "144 144\n",
      "CAM_BACK\n",
      "127 127\n",
      "CAM_BACK\n",
      "84 84\n",
      "CAM_BACK\n",
      "63 63\n",
      "CAM_BACK\n",
      "59 59\n",
      "290\n",
      "3 3\n",
      "291\n",
      "2 2\n",
      "3 3\n",
      "292\n",
      "3 3\n",
      "2 2\n",
      "3 3\n",
      "293\n",
      "2 2\n",
      "3 3\n",
      "2 2\n",
      "3 3\n",
      "294\n",
      "0 0\n",
      "2 2\n",
      "3 3\n",
      "2 2\n",
      "3 3\n",
      "295\n",
      "0 0\n",
      "0 0\n",
      "2 2\n",
      "3 3\n",
      "2 2\n",
      "3 3\n",
      "296\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "2 2\n",
      "3 3\n",
      "2 2\n",
      "297\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "2 2\n",
      "3 3\n",
      "298\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "2 2\n",
      "299\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "300\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "301\n",
      "2 2\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "302\n",
      "2 2\n",
      "2 2\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "303\n",
      "0 0\n",
      "2 2\n",
      "2 2\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "304\n",
      "2 2\n",
      "0 0\n",
      "2 2\n",
      "2 2\n",
      "0 0\n",
      "0 0\n",
      "305\n",
      "4 4\n",
      "2 2\n",
      "0 0\n",
      "2 2\n",
      "2 2\n",
      "0 0\n",
      "306\n",
      "0 0\n",
      "4 4\n",
      "2 2\n",
      "0 0\n",
      "2 2\n",
      "2 2\n",
      "307\n",
      "3 3\n",
      "0 0\n",
      "4 4\n",
      "2 2\n",
      "0 0\n",
      "2 2\n",
      "308\n",
      "3 3\n",
      "3 3\n",
      "0 0\n",
      "4 4\n",
      "2 2\n",
      "0 0\n",
      "309\n",
      "4 4\n",
      "3 3\n",
      "3 3\n",
      "0 0\n",
      "4 4\n",
      "2 2\n",
      "310\n",
      "3 3\n",
      "4 4\n",
      "3 3\n",
      "3 3\n",
      "0 0\n",
      "4 4\n",
      "311\n",
      "4 4\n",
      "3 3\n",
      "4 4\n",
      "3 3\n",
      "3 3\n",
      "0 0\n",
      "312\n",
      "1 1\n",
      "4 4\n",
      "3 3\n",
      "4 4\n",
      "3 3\n",
      "3 3\n",
      "313\n",
      "4 4\n",
      "1 1\n",
      "4 4\n",
      "3 3\n",
      "4 4\n",
      "3 3\n",
      "314\n",
      "0 0\n",
      "4 4\n",
      "1 1\n",
      "4 4\n",
      "3 3\n",
      "4 4\n",
      "315\n",
      "3 3\n",
      "316\n",
      "CAM_FRONT_RIGHT\n",
      "5 5\n",
      "3 3\n",
      "317\n",
      "CAM_FRONT_RIGHT\n",
      "4 4\n",
      "CAM_FRONT_RIGHT\n",
      "5 5\n",
      "3 3\n",
      "318\n",
      "CAM_FRONT_RIGHT\n",
      "3 3\n",
      "CAM_FRONT_RIGHT\n",
      "4 4\n",
      "CAM_FRONT_RIGHT\n",
      "5 5\n",
      "3 3\n",
      "319\n",
      "CAM_FRONT_RIGHT\n",
      "16 16\n",
      "CAM_FRONT_RIGHT\n",
      "3 3\n",
      "CAM_FRONT_RIGHT\n",
      "4 4\n",
      "CAM_FRONT_RIGHT\n",
      "5 5\n",
      "3 3\n",
      "320\n",
      "0 0\n",
      "321\n",
      "2 2\n",
      "0 0\n",
      "322\n",
      "1 1\n",
      "2 2\n",
      "0 0\n",
      "323\n",
      "0 0\n",
      "1 1\n",
      "2 2\n",
      "0 0\n",
      "324\n",
      "0 0\n",
      "0 0\n",
      "1 1\n",
      "2 2\n",
      "0 0\n",
      "325\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "1 1\n",
      "2 2\n",
      "0 0\n",
      "326\n",
      "8 8\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "1 1\n",
      "2 2\n",
      "327\n",
      "14 14\n",
      "8 8\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "1 1\n",
      "328\n",
      "10 10\n",
      "14 14\n",
      "8 8\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "329\n",
      "0 0\n",
      "330\n",
      "0 0\n",
      "0 0\n",
      "331\n",
      "1 1\n",
      "0 0\n",
      "0 0\n",
      "332\n",
      "1 1\n",
      "1 1\n",
      "0 0\n",
      "0 0\n",
      "333\n",
      "1 1\n",
      "1 1\n",
      "1 1\n",
      "0 0\n",
      "0 0\n",
      "334\n",
      "3 3\n",
      "1 1\n",
      "1 1\n",
      "1 1\n",
      "0 0\n",
      "0 0\n",
      "335\n",
      "0 0\n",
      "3 3\n",
      "1 1\n",
      "1 1\n",
      "1 1\n",
      "0 0\n",
      "336\n",
      "1 1\n",
      "0 0\n",
      "3 3\n",
      "1 1\n",
      "1 1\n",
      "1 1\n",
      "337\n",
      "1 1\n",
      "1 1\n",
      "0 0\n",
      "3 3\n",
      "1 1\n",
      "1 1\n",
      "338\n",
      "3 3\n",
      "1 1\n",
      "1 1\n",
      "0 0\n",
      "3 3\n",
      "1 1\n",
      "339\n",
      "1 1\n",
      "3 3\n",
      "1 1\n",
      "1 1\n",
      "0 0\n",
      "3 3\n",
      "340\n",
      "0 0\n",
      "1 1\n",
      "3 3\n",
      "1 1\n",
      "1 1\n",
      "0 0\n",
      "341\n",
      "1 1\n",
      "0 0\n",
      "1 1\n",
      "3 3\n",
      "1 1\n",
      "1 1\n",
      "342\n",
      "1 1\n",
      "1 1\n",
      "0 0\n",
      "1 1\n",
      "3 3\n",
      "1 1\n",
      "343\n",
      "0 0\n",
      "1 1\n",
      "1 1\n",
      "0 0\n",
      "1 1\n",
      "3 3\n",
      "344\n",
      "0 0\n",
      "0 0\n",
      "1 1\n",
      "1 1\n",
      "0 0\n",
      "1 1\n",
      "345\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "1 1\n",
      "1 1\n",
      "0 0\n",
      "346\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "1 1\n",
      "1 1\n",
      "347\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "1 1\n",
      "348\n",
      "1 1\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "349\n",
      "0 0\n",
      "1 1\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "350\n",
      "1 1\n",
      "0 0\n",
      "1 1\n",
      "0 0\n",
      "0 0\n",
      "0 0\n",
      "351\n",
      "0 0\n",
      "1 1\n",
      "0 0\n",
      "1 1\n",
      "0 0\n",
      "0 0\n",
      "352\n",
      "1 1\n",
      "0 0\n",
      "1 1\n",
      "0 0\n",
      "1 1\n",
      "0 0\n",
      "353\n",
      "0 0\n",
      "1 1\n",
      "0 0\n",
      "1 1\n",
      "0 0\n",
      "1 1\n",
      "354\n",
      "CAM_FRONT_RIGHT\n",
      "428 428\n",
      "355\n",
      "CAM_FRONT_RIGHT\n",
      "835 835\n",
      "CAM_FRONT_RIGHT\n",
      "428 428\n",
      "356\n",
      "CAM_BACK\n",
      "1 1\n",
      "357\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "1 1\n",
      "358\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "1 1\n",
      "359\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "1 1\n",
      "360\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "1 1\n",
      "361\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "1 1\n",
      "362\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "1 1\n",
      "363\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "1 1\n",
      "CAM_BACK\n",
      "3 3\n",
      "364\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "1 1\n",
      "365\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "0 0\n",
      "366\n",
      "CAM_BACK\n",
      "2 2\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "0 0\n",
      "367\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "2 2\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "3 3\n",
      "CAM_BACK\n",
      "0 0\n",
      "368\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "2 2\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "3 3\n",
      "369\n",
      "CAM_BACK\n",
      "2 2\n",
      "CAM_BACK\n",
      "0 0\n",
      "CAM_BACK\n",
      "4 4\n",
      "CAM_BACK\n",
      "2 2\n",
      "CAM_BACK\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-15-2c4113263c62>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m     33\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     34\u001b[0m         \u001b[0m_\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mboxes\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0m_\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnusc\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_sample_data\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlidar\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mselected_anntokens\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mselected_anntokens\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---> 35\u001b[0;31m         \u001b[0mmask\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mpoints_in_box\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mboxes\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpc\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpoints\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;36m3\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;31m#mask for lidar points of instance in sensor frame\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     36\u001b[0m         \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpc\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpoints\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mmask\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mannotation\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'num_lidar_pts'\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[1;32m     37\u001b[0m         \u001b[0;32massert\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpc\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpoints\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mmask\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0mannotation\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'num_lidar_pts'\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~/.local/lib/python3.7/site-packages/nuscenes/utils/geometry_utils.py\u001b[0m in \u001b[0;36mpoints_in_box\u001b[0;34m(box, points, wlh_factor)\u001b[0m\n\u001b[1;32m    135\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    136\u001b[0m     \u001b[0miv\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mv\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 137\u001b[0;31m     \u001b[0mjv\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mj\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mv\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    138\u001b[0m     \u001b[0mkv\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mk\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mv\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    139\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "count = 0 \n",
    "for i, token in enumerate(mini_train): \n",
    "    print(count)\n",
    "    instance_token, sample_token = token.split(\"_\")\n",
    "    helper.get_past_for_agent(instance_token, sample_token, seconds=3, in_agent_frame=False, just_xy=False)\n",
    "\n",
    "    for frame in helper.get_past_for_agent(instance_token, sample_token, seconds=3, in_agent_frame=False, just_xy=False):\n",
    "        \n",
    "        frame_sample_token, frame_instance_token = frame['sample_token'], frame['instance_token']\n",
    "        \n",
    "        sample = nusc.get('sample', frame_sample_token)\n",
    "        lidar_token = sample['data']['LIDAR_TOP']\n",
    "        annotation = helper.get_sample_annotation(frame_instance_token, frame_sample_token)\n",
    "        selected_anntokens = annotation['token']\n",
    "        \n",
    "\n",
    "        ann_record = nusc.get('sample_annotation', selected_anntokens)\n",
    "        sample_record = nusc.get('sample', ann_record['sample_token'])\n",
    "\n",
    "        # Figure out which camera the object is fully visible in (this may return nothing)\n",
    "        cams = [key for key in sample_record['data'].keys() if 'CAM' in key]\n",
    "        for cam in cams: \n",
    "            _, boxes, _ = nusc.get_sample_data(sample_record['data'][cam], box_vis_level=BoxVisibility.ANY,\n",
    "                                                    selected_anntokens=[selected_anntokens])\n",
    "            if len(boxes) > 0: \n",
    "                break\n",
    "        \n",
    "        if cam != 'CAM_FRONT':\n",
    "            print(cam)\n",
    "        cam = sample_record['data'][cam]\n",
    "        lidar = sample_record['data']['LIDAR_TOP']\n",
    "        pc, times = get_lidar_data_pc(lidar, nsweeps=1)        \n",
    "\n",
    "        _, boxes, _ = nusc.get_sample_data(lidar, selected_anntokens=[selected_anntokens])\n",
    "        mask = points_in_box(boxes[0], pc.points[:3,:]) #mask for lidar points of instance in sensor frame\n",
    "        print(pc.points[:,mask].shape[1], annotation['num_lidar_pts'])\n",
    "        assert(pc.points[:,mask].shape[1] == annotation['num_lidar_pts'])\n",
    "        \n",
    "    count+=1\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "bc38961ca0ac4b14ab90e547ba79fbb6 39586f9d59004284a7114a68825e8eec\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "1"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "instance_token, sample_token = mini_train[0].split(\"_\")\n",
    "print(instance_token, sample_token)\n",
    "annotation = helper.get_sample_annotation(instance_token, sample_token)\n",
    "# next_token = annotation['next']\n",
    "# annotation\n",
    "helper.get_past_for_agent(instance_token, sample_token, seconds=3, in_agent_frame=False, just_xy=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To get the future/past of an agent, use the `get_past_for_agent`/`get_future_for_agent` methods. If the `in_agent_frame` parameter is set to true, the coordinates will be in the agent's local coordinate frame. Otherwise, they will be in the global frame."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4, 2)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "future_xy_local = helper.get_past_for_agent(instance_token, sample_token, seconds=2, in_agent_frame=True)\n",
    "future_xy_local.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The agent's coordinate frame is centered on the agent's current location and the agent's heading is aligned with the positive y axis. For example, the last coordinate in `future_xy_local` corresponds to a location 0.31 meters to the left and 9.67 meters in front of the agents starting location."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 391.347, 1144.208],\n",
       "       [ 390.512, 1142.201],\n",
       "       [ 389.29 , 1139.46 ],\n",
       "       [ 388.168, 1136.675],\n",
       "       [ 386.878, 1133.783],\n",
       "       [ 385.638, 1131.14 ],\n",
       "       [ 384.005, 1128.317],\n",
       "       [ 382.188, 1125.617]])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "future_xy_global = helper.get_future_for_agent(instance_token, sample_token, seconds=4, in_agent_frame=False)\n",
    "future_xy_global"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note that you can also return the entire annotation record by passing `just_xy=False`. However in this case, `in_agent_frame` is not taken into account."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[{'token': '247a25c59f914adabee9460bd8307196',\n",
       "  'sample_token': 'f1e3d9d08f044c439ce86a2d6fcca57b',\n",
       "  'instance_token': 'bc38961ca0ac4b14ab90e547ba79fbb6',\n",
       "  'visibility_token': '4',\n",
       "  'attribute_tokens': ['cb5118da1ab342aa947717dc53544259'],\n",
       "  'translation': [391.347, 1144.208, 0.841],\n",
       "  'size': [1.708, 4.01, 1.631],\n",
       "  'rotation': [-0.5443682117180475, 0.0, 0.0, 0.8388463804957943],\n",
       "  'prev': '7bcf4bc87bf143588230254034eada12',\n",
       "  'next': 'e25b9e7019814d53876ff2697df7a2de',\n",
       "  'num_lidar_pts': 20,\n",
       "  'num_radar_pts': 4,\n",
       "  'category_name': 'vehicle.car'},\n",
       " {'token': 'e25b9e7019814d53876ff2697df7a2de',\n",
       "  'sample_token': '4f545737bf3347fbbc9af60b0be9a963',\n",
       "  'instance_token': 'bc38961ca0ac4b14ab90e547ba79fbb6',\n",
       "  'visibility_token': '4',\n",
       "  'attribute_tokens': ['cb5118da1ab342aa947717dc53544259'],\n",
       "  'translation': [390.512, 1142.201, 0.891],\n",
       "  'size': [1.708, 4.01, 1.631],\n",
       "  'rotation': [-0.5443682117180475, 0.0, 0.0, 0.8388463804957943],\n",
       "  'prev': '247a25c59f914adabee9460bd8307196',\n",
       "  'next': 'fe33c018573e4abda3ff8de0566ee800',\n",
       "  'num_lidar_pts': 24,\n",
       "  'num_radar_pts': 2,\n",
       "  'category_name': 'vehicle.car'},\n",
       " {'token': 'fe33c018573e4abda3ff8de0566ee800',\n",
       "  'sample_token': '7626dde27d604ac28a0240bdd54eba7a',\n",
       "  'instance_token': 'bc38961ca0ac4b14ab90e547ba79fbb6',\n",
       "  'visibility_token': '4',\n",
       "  'attribute_tokens': ['cb5118da1ab342aa947717dc53544259'],\n",
       "  'translation': [389.29, 1139.46, 0.941],\n",
       "  'size': [1.708, 4.01, 1.631],\n",
       "  'rotation': [-0.5443682117180475, 0.0, 0.0, 0.8388463804957943],\n",
       "  'prev': 'e25b9e7019814d53876ff2697df7a2de',\n",
       "  'next': '2c1a8ae13d76498c838a1fb733ff8700',\n",
       "  'num_lidar_pts': 30,\n",
       "  'num_radar_pts': 2,\n",
       "  'category_name': 'vehicle.car'},\n",
       " {'token': '2c1a8ae13d76498c838a1fb733ff8700',\n",
       "  'sample_token': 'be99ffc878b24aca8956bbb4e0f97d0c',\n",
       "  'instance_token': 'bc38961ca0ac4b14ab90e547ba79fbb6',\n",
       "  'visibility_token': '4',\n",
       "  'attribute_tokens': ['cb5118da1ab342aa947717dc53544259'],\n",
       "  'translation': [388.168, 1136.675, 0.991],\n",
       "  'size': [1.708, 4.01, 1.631],\n",
       "  'rotation': [-0.5443682117180475, 0.0, 0.0, 0.8388463804957943],\n",
       "  'prev': 'fe33c018573e4abda3ff8de0566ee800',\n",
       "  'next': '2064ecd1630a428184fb8715b226bf9b',\n",
       "  'num_lidar_pts': 35,\n",
       "  'num_radar_pts': 4,\n",
       "  'category_name': 'vehicle.car'},\n",
       " {'token': '2064ecd1630a428184fb8715b226bf9b',\n",
       "  'sample_token': '9813c23a5f1448b09bb7910fea9baf20',\n",
       "  'instance_token': 'bc38961ca0ac4b14ab90e547ba79fbb6',\n",
       "  'visibility_token': '4',\n",
       "  'attribute_tokens': ['cb5118da1ab342aa947717dc53544259'],\n",
       "  'translation': [386.878, 1133.783, 0.991],\n",
       "  'size': [1.708, 4.01, 1.631],\n",
       "  'rotation': [-0.5443682117180475, 0.0, 0.0, 0.8388463804957943],\n",
       "  'prev': '2c1a8ae13d76498c838a1fb733ff8700',\n",
       "  'next': 'ec5d36663d994871be2aa33248629244',\n",
       "  'num_lidar_pts': 33,\n",
       "  'num_radar_pts': 3,\n",
       "  'category_name': 'vehicle.car'},\n",
       " {'token': 'ec5d36663d994871be2aa33248629244',\n",
       "  'sample_token': '023c4df2d451409881d8e6ea82f14704',\n",
       "  'instance_token': 'bc38961ca0ac4b14ab90e547ba79fbb6',\n",
       "  'visibility_token': '4',\n",
       "  'attribute_tokens': ['cb5118da1ab342aa947717dc53544259'],\n",
       "  'translation': [385.638, 1131.14, 1.041],\n",
       "  'size': [1.708, 4.01, 1.631],\n",
       "  'rotation': [-0.5443682117180475, 0.0, 0.0, 0.8388463804957943],\n",
       "  'prev': '2064ecd1630a428184fb8715b226bf9b',\n",
       "  'next': '966ef9fb97bc46d98ef2a4e999c37e86',\n",
       "  'num_lidar_pts': 31,\n",
       "  'num_radar_pts': 3,\n",
       "  'category_name': 'vehicle.car'}]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "helper.get_future_for_agent(instance_token, sample_token, seconds=3, in_agent_frame=True, just_xy=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[{'token': 'b2b43ef63f5242b2a4c0b794e673782d',\n",
       "  'sample_token': 'e0845f5322254dafadbbed75aaa07969',\n",
       "  'instance_token': 'bc38961ca0ac4b14ab90e547ba79fbb6',\n",
       "  'visibility_token': '4',\n",
       "  'attribute_tokens': ['cb5118da1ab342aa947717dc53544259'],\n",
       "  'translation': [392.641, 1147.242, 0.816],\n",
       "  'size': [1.708, 4.01, 1.631],\n",
       "  'rotation': [-0.5443682117180475, 0.0, 0.0, 0.8388463804957943],\n",
       "  'prev': 'b41e15b89fd44709b439de95dd723617',\n",
       "  'next': '7bcf4bc87bf143588230254034eada12',\n",
       "  'num_lidar_pts': 13,\n",
       "  'num_radar_pts': 3,\n",
       "  'category_name': 'vehicle.car'},\n",
       " {'token': 'b41e15b89fd44709b439de95dd723617',\n",
       "  'sample_token': '356d81f38dd9473ba590f39e266f54e5',\n",
       "  'instance_token': 'bc38961ca0ac4b14ab90e547ba79fbb6',\n",
       "  'visibility_token': '4',\n",
       "  'attribute_tokens': ['cb5118da1ab342aa947717dc53544259'],\n",
       "  'translation': [392.836, 1148.208, 0.791],\n",
       "  'size': [1.708, 4.01, 1.631],\n",
       "  'rotation': [-0.5443682117180475, 0.0, 0.0, 0.8388463804957943],\n",
       "  'prev': 'a286c9633fa34da5b978758f348996b0',\n",
       "  'next': 'b2b43ef63f5242b2a4c0b794e673782d',\n",
       "  'num_lidar_pts': 10,\n",
       "  'num_radar_pts': 2,\n",
       "  'category_name': 'vehicle.car'},\n",
       " {'token': 'a286c9633fa34da5b978758f348996b0',\n",
       "  'sample_token': '39586f9d59004284a7114a68825e8eec',\n",
       "  'instance_token': 'bc38961ca0ac4b14ab90e547ba79fbb6',\n",
       "  'visibility_token': '4',\n",
       "  'attribute_tokens': ['cb5118da1ab342aa947717dc53544259'],\n",
       "  'translation': [392.945, 1148.426, 0.766],\n",
       "  'size': [1.708, 4.01, 1.631],\n",
       "  'rotation': [-0.5443682117180475, 0.0, 0.0, 0.8388463804957943],\n",
       "  'prev': '16140fbf143d4e26a4a7613cbd3aa0e8',\n",
       "  'next': 'b41e15b89fd44709b439de95dd723617',\n",
       "  'num_lidar_pts': 0,\n",
       "  'num_radar_pts': 0,\n",
       "  'category_name': 'vehicle.car'},\n",
       " {'token': '16140fbf143d4e26a4a7613cbd3aa0e8',\n",
       "  'sample_token': 'ca9a282c9e77460f8360f564131a8af5',\n",
       "  'instance_token': 'bc38961ca0ac4b14ab90e547ba79fbb6',\n",
       "  'visibility_token': '4',\n",
       "  'attribute_tokens': ['cb5118da1ab342aa947717dc53544259'],\n",
       "  'translation': [393.357, 1149.173, 0.741],\n",
       "  'size': [1.708, 4.01, 1.631],\n",
       "  'rotation': [-0.5443682117180475, 0.0, 0.0, 0.8388463804957943],\n",
       "  'prev': '',\n",
       "  'next': 'a286c9633fa34da5b978758f348996b0',\n",
       "  'num_lidar_pts': 4,\n",
       "  'num_radar_pts': 2,\n",
       "  'category_name': 'vehicle.car'}]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "helper.get_past_for_agent(instance_token, sample_token, seconds=3, in_agent_frame=False, just_xy=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If you would like to return the data for the entire sample, as opposed to one agent in the sample, you can use the `get_annotations_for_sample` method. This will return a list of records for each annotated agent in the sample."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "78"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample = helper.get_annotations_for_sample(sample_token)\n",
    "len(sample)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note that there are `get_future_for_sample` and `get_past_for_sample` methods that are analogous to the `get_future_for_agent` and `get_past_for_agent` methods.\n",
    "\n",
    "We also provide methods to compute the velocity, acceleration, and heading change rate of an agent at a given point in time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Velocity: 4.385040264738063\n",
      "\n",
      "Acceleration: 0.30576530453207523\n",
      "\n",
      "Heading Change Rate: 0.0\n"
     ]
    }
   ],
   "source": [
    "# We get new instance and sample tokens because these methods require computing the difference between records.\n",
    "instance_token_2, sample_token_2 = mini_train[5].split(\"_\")\n",
    "\n",
    "# Meters / second.\n",
    "print(f\"Velocity: {helper.get_velocity_for_agent(instance_token_2, sample_token_2)}\\n\")\n",
    "\n",
    "# Meters / second^2.\n",
    "print(f\"Acceleration: {helper.get_acceleration_for_agent(instance_token_2, sample_token_2)}\\n\")\n",
    "\n",
    "# Radians / second.\n",
    "print(f\"Heading Change Rate: {helper.get_heading_change_rate_for_agent(instance_token_2, sample_token_2)}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Changes to the Map API\n",
    "\n",
    "We've added a couple of methods to the Map API to help query lane center line information."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "from nuscenes.map_expansion.map_api import NuScenesMap\n",
    "nusc_map = NuScenesMap(map_name='singapore-onenorth', dataroot=DATAROOT)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To get the closest lane to a location, use the `get_closest_lane` method. To see the internal data representation of the lane, use the `get_lane_record` method. \n",
    "You can also explore the connectivity of the lanes, with the `get_outgoing_lanes` and `get_incoming_lane` methods."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'5933500a-f0f2-4d69-9bbc-83b875e4a73e'"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x, y, yaw = 395, 1095, 0\n",
    "closest_lane = nusc_map.get_closest_lane(x, y, radius=2)\n",
    "closest_lane"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[{'start_pose': [421.2419602954602, 1087.9127960414617, 2.739593514975998],\n",
       "  'end_pose': [391.7142849867393, 1100.464077182952, 2.7365754617298705],\n",
       "  'shape': 'LSR',\n",
       "  'radius': 999.999,\n",
       "  'segment_length': [0.23651121617864976,\n",
       "   28.593481378991886,\n",
       "   3.254561444252876]}]"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lane_record = nusc_map.get_lane(closest_lane)\n",
    "lane_record"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['f24a067b-d650-47d0-8664-039d648d7c0d']"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nusc_map.get_incoming_lane_ids(closest_lane)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['0282d0e3-b6bf-4bcd-be24-35c9ce4c6591',\n",
       " '28d15254-0ef9-48c3-9e06-dc5a25b31127']"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nusc_map.get_outgoing_lane_ids(closest_lane)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To help manipulate the lanes, we've added an `arcline_path_utils` module. For example, something you might want to do is discretize a lane into a sequence of poses."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(421.2419602954602, 1087.9127960414617, 2.739593514975998),\n",
       " (420.34712994585345, 1088.2930152148274, 2.739830026428688),\n",
       " (419.45228865726136, 1088.6732086473173, 2.739830026428688),\n",
       " (418.5574473686693, 1089.0534020798073, 2.739830026428688),\n",
       " (417.66260608007724, 1089.433595512297, 2.739830026428688),\n",
       " (416.76776479148515, 1089.813788944787, 2.739830026428688),\n",
       " (415.87292350289306, 1090.1939823772768, 2.739830026428688),\n",
       " (414.97808221430097, 1090.5741758097668, 2.739830026428688),\n",
       " (414.0832409257089, 1090.9543692422567, 2.739830026428688),\n",
       " (413.1883996371168, 1091.3345626747464, 2.739830026428688),\n",
       " (412.29355834852475, 1091.7147561072363, 2.739830026428688),\n",
       " (411.39871705993266, 1092.0949495397263, 2.739830026428688),\n",
       " (410.5038757713406, 1092.4751429722162, 2.739830026428688),\n",
       " (409.6090344827485, 1092.8553364047061, 2.739830026428688),\n",
       " (408.7141931941564, 1093.2355298371958, 2.739830026428688),\n",
       " (407.81935190556436, 1093.6157232696858, 2.739830026428688),\n",
       " (406.92451061697227, 1093.9959167021757, 2.739830026428688),\n",
       " (406.0296693283802, 1094.3761101346656, 2.739830026428688),\n",
       " (405.1348280397881, 1094.7563035671556, 2.739830026428688),\n",
       " (404.239986751196, 1095.1364969996453, 2.739830026428688),\n",
       " (403.3451454626039, 1095.5166904321352, 2.739830026428688),\n",
       " (402.4503041740119, 1095.8968838646251, 2.739830026428688),\n",
       " (401.5554628854198, 1096.277077297115, 2.739830026428688),\n",
       " (400.6606215968277, 1096.657270729605, 2.739830026428688),\n",
       " (399.7657803082356, 1097.0374641620947, 2.739830026428688),\n",
       " (398.8709390196435, 1097.4176575945846, 2.739830026428688),\n",
       " (397.9760977310515, 1097.7978510270746, 2.739830026428688),\n",
       " (397.0812564424594, 1098.1780444595645, 2.739830026428688),\n",
       " (396.1864151538673, 1098.5582378920544, 2.739830026428688),\n",
       " (395.2915738652752, 1098.9384313245444, 2.739830026428688),\n",
       " (394.3967548911081, 1099.318677260896, 2.739492242286598),\n",
       " (393.5022271882191, 1099.69960782173, 2.738519982101022),\n",
       " (392.60807027168346, 1100.0814079160527, 2.737547721915446),\n",
       " (391.71428498673856, 1100.4640771829522, 2.7365754617298705)]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from nuscenes.map_expansion import arcline_path_utils\n",
    "poses = arcline_path_utils.discretize_lane(lane_record, resolution_meters=1)\n",
    "poses"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Given a query pose, you can also find the closest pose on a lane."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "closest_pose_on_lane, distance_along_lane = arcline_path_utils.project_pose_to_lane((x, y, yaw), lane_record)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "395 1095 0\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(396.25524909914367, 1098.5289922434013, 2.739830026428688)"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(x, y, yaw)\n",
    "closest_pose_on_lane"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "27.5"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Meters.\n",
    "distance_along_lane"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To find the entire length of the lane, you can use the `length_of_lane` function."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "32.08455403942341"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "arcline_path_utils.length_of_lane(lane_record)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can also compute the curvature of a lane at a given distance along the lane."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 0 means it is a straight lane.\n",
    "arcline_path_utils.get_curvature_at_distance_along_lane(distance_along_lane, lane_record)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. Input Representation\n",
    "\n",
    "It is common in the prediction literature to represent the state of an agent as a tensor containing information about the semantic map (such as the drivable area and walkways), as well the past locations of surrounding agents.\n",
    "\n",
    "Each paper in the field chooses to represent the input in a slightly different way. For example, [CoverNet](https://arxiv.org/pdf/1911.10298.pdf) and [MTP](https://arxiv.org/pdf/1808.05819.pdf) choose to rasterize the map information and agent locations into a three channel RGB image. But [Rules of the Road](http://openaccess.thecvf.com/content_CVPR_2019/papers/Hong_Rules_of_the_Road_Predicting_Driving_Behavior_With_a_Convolutional_CVPR_2019_paper.pdf) decides to use a \"taller\" tensor with information represented in different channels.\n",
    "\n",
    "We provide a module called `input_representation` that is meant to make it easy for you to define your own input representation. In short, you need to define your own `StaticLayerRepresentation`, `AgentRepresentation`, and `Combinator`.\n",
    "\n",
    "The `StaticLayerRepresentation` controls how the static map information is represented. The `AgentRepresentation` controls how the locations of the agents in the scene are represented. The `Combinator` controls how these two sources of information are combined into a single tensor.\n",
    "\n",
    "For more information, consult `input_representation/interface.py`.\n",
    "\n",
    "To help get you started, we've provided implementations of input representation used in CoverNet and MTP."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "static_layers.py - Loading Map: singapore-queenstown\n",
      "static_layers.py - Loading Map: singapore-onenorth\n",
      "static_layers.py - Loading Map: boston-seaport\n",
      "static_layers.py - Loading Map: singapore-hollandvillage\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7fa210c6bf60>"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "from nuscenes.prediction.input_representation.static_layers import StaticLayerRasterizer\n",
    "from nuscenes.prediction.input_representation.agents import AgentBoxesWithFadedHistory\n",
    "from nuscenes.prediction.input_representation.interface import InputRepresentation\n",
    "from nuscenes.prediction.input_representation.combinators import Rasterizer\n",
    "\n",
    "static_layer_rasterizer = StaticLayerRasterizer(helper)\n",
    "agent_rasterizer = AgentBoxesWithFadedHistory(helper, seconds_of_history=1)\n",
    "mtp_input_representation = InputRepresentation(static_layer_rasterizer, agent_rasterizer, Rasterizer())\n",
    "\n",
    "instance_token_img, sample_token_img = 'bc38961ca0ac4b14ab90e547ba79fbb6', '7626dde27d604ac28a0240bdd54eba7a'\n",
    "anns = [ann for ann in nusc.sample_annotation if ann['instance_token'] == instance_token_img]\n",
    "img = mtp_input_representation.make_input_representation(instance_token_img, sample_token_img)\n",
    "\n",
    "plt.imshow(img)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model Implementations\n",
    "\n",
    "We've provided PyTorch implementations for CoverNet and MTP. Below we show, how to make predictions on the previously created input representation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "from nuscenes.prediction.models.backbone import ResNetBackbone\n",
    "from nuscenes.prediction.models.mtp import MTP\n",
    "from nuscenes.prediction.models.covernet import CoverNet\n",
    "import torch"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Both models take a CNN backbone as a parameter. We've provided wrappers for ResNet and MobileNet v2. In this example, we'll use ResNet50."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "backbone = ResNetBackbone('resnet50')\n",
    "mtp = MTP(backbone, num_modes=2)\n",
    "\n",
    "# Note that the value of num_modes depends on the size of the lattice used for CoverNet.\n",
    "covernet = CoverNet(backbone, num_modes=64)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The second input is a tensor containing the velocity, acceleration, and heading change rate for the agent."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "agent_state_vector = torch.Tensor([[helper.get_velocity_for_agent(instance_token_img, sample_token_img),\n",
    "                                    helper.get_acceleration_for_agent(instance_token_img, sample_token_img),\n",
    "                                    helper.get_heading_change_rate_for_agent(instance_token_img, sample_token_img)]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([1, 3])"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "agent_state_vector.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "image_tensor = torch.Tensor(img).permute(2, 0, 1).unsqueeze(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([1, 3, 500, 500])"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "image_tensor.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([1, 50])"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Output has 50 entries.\n",
    "# The first 24 are x,y coordinates (in the agent frame) over the next 6 seconds at 2 Hz for the first mode.\n",
    "# The second 24 are the x,y coordinates for the second mode.\n",
    "# The last 2 are the logits of the mode probabilities\n",
    "mtp(image_tensor, agent_state_vector).shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([1, 64])\n"
     ]
    }
   ],
   "source": [
    "# CoverNet outputs a probability distribution over the trajectory set.\n",
    "# These are the logits of the probabilities\n",
    "logits = covernet(image_tensor, agent_state_vector)\n",
    "print(logits.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The CoverNet model outputs a probability distribution over a set of trajectories. To be able to interpret the predictions, and perform inference with CoverNet, you need to download the trajectory sets from the nuscenes website. Download them from this [link](https://www.nuscenes.org/public/nuscenes-prediction-challenge-trajectory-sets.zip) and unzip them in a directory of your choice.\n",
    "\n",
    "Uncomment the following code when you do so:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[[[-1.2467e-02,  1.3043e+00],\n",
       "          [-2.4934e-02,  2.6086e+00],\n",
       "          [-2.9038e-02,  4.9068e+00],\n",
       "          ...,\n",
       "          [-2.9913e-01, -5.2803e-01],\n",
       "          [-3.4498e-01, -3.7897e+00],\n",
       "          [-3.7626e-01, -4.0333e+00]],\n",
       "\n",
       "         [[-2.9563e-01,  3.4745e+00],\n",
       "          [-9.9253e-01,  6.8860e+00],\n",
       "          [-2.0803e+00,  1.0195e+01],\n",
       "          ...,\n",
       "          [-1.8827e+01,  3.2410e+01],\n",
       "          [-2.1741e+01,  3.5716e+01],\n",
       "          [-2.5389e+01,  4.0019e+01]],\n",
       "\n",
       "         [[-2.1105e-02,  4.3084e+00],\n",
       "          [-3.1000e-02,  8.5625e+00],\n",
       "          [-1.8631e-02,  1.2780e+01],\n",
       "          ...,\n",
       "          [ 6.5467e-02,  4.3218e+01],\n",
       "          [ 4.3305e-02,  4.7518e+01],\n",
       "          [ 1.7727e-02,  5.2321e+01]],\n",
       "\n",
       "         ...,\n",
       "\n",
       "         [[-2.4342e-02,  6.0260e+00],\n",
       "          [-1.0404e-01,  1.1295e+01],\n",
       "          [-9.2170e-02,  1.7083e+01],\n",
       "          ...,\n",
       "          [ 3.6829e-01,  5.7070e+01],\n",
       "          [ 2.6024e-01,  6.3046e+01],\n",
       "          [ 7.5568e-02,  6.9022e+01]],\n",
       "\n",
       "         [[-2.9332e-01,  4.5648e+00],\n",
       "          [-1.1342e+00,  9.8443e+00],\n",
       "          [-2.6300e+00,  1.6081e+01],\n",
       "          ...,\n",
       "          [-1.4511e+01,  5.7503e+01],\n",
       "          [-1.6235e+01,  6.3654e+01],\n",
       "          [-1.7946e+01,  6.9744e+01]],\n",
       "\n",
       "         [[ 5.1619e-04,  3.5624e+00],\n",
       "          [ 1.0324e-03,  7.1249e+00],\n",
       "          [-2.1119e-02,  9.4521e+00],\n",
       "          ...,\n",
       "          [ 6.3271e-03,  2.9263e+01],\n",
       "          [ 4.2924e-02,  3.2302e+01],\n",
       "          [ 3.6384e-02,  3.5357e+01]]]])"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pickle\n",
    "\n",
    "# Epsilon is the amount of coverage in the set, \n",
    "# i.e. a real world trajectory is at most 8 meters from a trajectory in this set\n",
    "# We released the set for epsilon = 2, 4, 8. Consult the paper for more information\n",
    "# on how this set was created\n",
    "\n",
    "PATH_TO_EPSILON_8_SET = \"/home/armahade/ECE285/AV/final_proj/data/sets/nuscenes/nuscenes-prediction-challenge-trajectory-sets/epsilon_8.pkl\"\n",
    "trajectories = pickle.load(open(PATH_TO_EPSILON_8_SET, 'rb'))\n",
    "\n",
    "# Saved them as a list of lists\n",
    "trajectories = torch.Tensor(trajectories)\n",
    "\n",
    "# Print 5 most likely predictions\n",
    "trajectories[logits.argsort(descending=True)[:5]]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We also provide two physics-based models - A constant velocity and heading model and a physics oracle. The physics oracle estimates the future trajectory of the agent with several physics based models and chooses the one that is closest to the ground truth. It represents the best performance a purely physics based model could achieve on the dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "from nuscenes.prediction.models.physics import ConstantVelocityHeading, PhysicsOracle\n",
    "\n",
    "cv_model = ConstantVelocityHeading(sec_from_now=6, helper=helper)\n",
    "physics_oracle = PhysicsOracle(sec_from_now=6, helper=helper)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The physics models can be called as functions. They take as input a string of the instance and sample token of the agent concatenated with an underscore (\"_\").\n",
    "\n",
    "The output is a `Prediction` data type. The `Prediction` data type stores the predicted trajectories and their associated probabilities for the agent. We'll go over the `Prediction` type in greater detail in the next section."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Prediction(instance=bc38961ca0ac4b14ab90e547ba79fbb6, sample=7626dde27d604ac28a0240bdd54eba7a, prediction=[[[ 388.17909232 1136.96919062]\n",
       "  [ 387.06818463 1134.47838124]\n",
       "  [ 385.95727695 1131.98757186]\n",
       "  [ 384.84636926 1129.49676248]\n",
       "  [ 383.73546158 1127.0059531 ]\n",
       "  [ 382.6245539  1124.51514372]\n",
       "  [ 381.51364621 1122.02433435]\n",
       "  [ 380.40273853 1119.53352497]\n",
       "  [ 379.29183084 1117.04271559]\n",
       "  [ 378.18092316 1114.55190621]\n",
       "  [ 377.07001548 1112.06109683]\n",
       "  [ 375.95910779 1109.57028745]]], probabilities=[1])"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv_model(f\"{instance_token_img}_{sample_token_img}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Prediction(instance=bc38961ca0ac4b14ab90e547ba79fbb6, sample=7626dde27d604ac28a0240bdd54eba7a, prediction=[[[ 388.17909232 1136.96919062]\n",
       "  [ 387.06818463 1134.47838124]\n",
       "  [ 385.95727695 1131.98757186]\n",
       "  [ 384.84636926 1129.49676248]\n",
       "  [ 383.73546158 1127.0059531 ]\n",
       "  [ 382.6245539  1124.51514372]\n",
       "  [ 381.51364621 1122.02433435]\n",
       "  [ 380.40273853 1119.53352497]\n",
       "  [ 379.29183084 1117.04271559]\n",
       "  [ 378.18092316 1114.55190621]\n",
       "  [ 377.07001548 1112.06109683]\n",
       "  [ 375.95910779 1109.57028745]]], probabilities=[1])"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "physics_oracle(f\"{instance_token_img}_{sample_token_img}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6. Making a submission to the challenge\n",
    "\n",
    "Participants must submit a zipped json file containing serialized `Predictions` for each agent in the validation set.\n",
    "\n",
    "The previous section introduced the `Prediction` data type. In this section, we explain the format in greater detail. \n",
    "\n",
    "A `Prediction` consists of four fields:\n",
    "\n",
    "1. instance: The instance token for the agent.\n",
    "2. sample: The sample token for the agent.\n",
    "3. prediction: Prediction from model. A prediction can consist of up to 25 proposed trajectories. This field must be a numpy array with three dimensions (number of trajectories (also called modes), number of timesteps, 2).\n",
    "4. probabilities: The probability corresponding to each predicted mode. This is a numpy array with shape `(number_of_modes,)`.\n",
    "\n",
    "You will get an error if any of these conditions are violated."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from nuscenes.eval.prediction.data_classes import Prediction\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# This would raise an error because instance is not a string.\n",
    "\n",
    "#Prediction(instance=1, sample=sample_token_img,\n",
    "#           prediction=np.ones((1, 12, 2)), probabilities=np.array([1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# This would raise an error because sample is not a string.\n",
    "\n",
    "#Prediction(instance=instance_token_img, sample=2,\n",
    "#           prediction=np.ones((1, 12, 2)), probabilities=np.array([1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# This would raise an error because prediction is not a numpy array.\n",
    "\n",
    "#Prediction(instance=instance_token_img, sample=sample_token_img,\n",
    "#           prediction=np.ones((1, 12, 2)).tolist(), probabilities=np.array([1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# This would throw an error because probabilities is not a numpy array. Uncomment to see.\n",
    "\n",
    "#Prediction(instance=instance_token_img, sample=sample_token_img,\n",
    "#           prediction=np.ones((1, 12, 2)), probabilities=[0.3])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# This would throw an error because there are more than 25 predicted modes. Uncomment to see.\n",
    "\n",
    "#Prediction(instance=instance_token_img, sample=sample_token_img,\n",
    "#           prediction=np.ones((30, 12, 2)), probabilities=np.array([1/30]*30))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# This would throw an error because the number of predictions and probabilities don't match. Uncomment to see.\n",
    "\n",
    "#Prediction(instance=instance_token_img, sample=sample_token_img,\n",
    "           #prediction=np.ones((13, 12, 2)), probabilities=np.array([1/12]*12))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To make a submission to the challenge, store your model predictions in a python list and save it to json. Then, upload a zipped version of your file to the eval server. \n",
    "\n",
    "For an example, see `eval/prediction/baseline_model_inference.py`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}