{
  "id": 340630,
  "title": "Question for top-teams:  are your solutions real-time?",
  "url": "/competitions/smartphone-decimeter-2022/discussion/340630",
  "author_name": "",
  "post_date": "2022-07-30T06:34:10.873071Z",
  "votes": 4,
  "comment_count": 2,
  "views": 0,
  "content": "<p>May I ask if your solutions are real-time localization or need post-processing? One of my recent projects requires real-time precise localization, but I followed some public open-source solutions and find most solutions require whole route information in this competition (i.e., access to all data after route completion) and then do filtering. I appreciate it if you can help.</p>",
  "messages": [
    {
      "id": "1876887",
      "postDate": "07/30/2022 06:34:10",
      "content": "<p>May I ask if your solutions are real-time localization or need post-processing? One of my recent projects requires real-time precise localization, but I followed some public open-source solutions and find most solutions require whole route information in this competition (i.e., access to all data after route completion) and then do filtering. I appreciate it if you can help.</p>",
      "rawMarkdown": "May I ask if your solutions are real-time localization or need post-processing? One of my recent projects requires real-time precise localization, but I followed some public open-source solutions and find most solutions require whole route information in this competition (i.e., access to all data after route completion) and then do filtering. I appreciate it if you can help.",
      "votes": null
    },
    {
      "id": "1877267",
      "postDate": "07/30/2022 13:21:23",
      "content": "<p>Many people, myself included, are doing Kalman Smoothing, which effectively runs a Kalman filter forward, and then  backward over the results, combining the results appropriately.   This means that future data is used for positioning, and thus is not a real time method.</p>",
      "rawMarkdown": "Many people, myself included, are doing Kalman Smoothing, which effectively runs a Kalman filter forward, and then  backward over the results, combining the results appropriately.   This means that future data is used for positioning, and thus is not a real time method.",
      "votes": null
    },
    {
      "id": "1877471",
      "postDate": "07/30/2022 17:05:12",
      "content": "<p>Got it and thanks for the comment. I am wondering if it is possible to do real-time localization since I think that's more like our daily routine. <a href=\"https://www.kaggle.com/mohammedkhider\" target=\"_blank\">@mohammedkhider</a> Will you hold the same type of competition for next year? Do you think we can do real-time localization down to the decimeter? Thanks!</p>",
      "rawMarkdown": "Got it and thanks for the comment. I am wondering if it is possible to do real-time localization since I think that's more like our daily routine. @mohammedkhider Will you hold the same type of competition for next year? Do you think we can do real-time localization down to the decimeter? Thanks!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1877267,
      "author_name": "solverworld",
      "author_url": "",
      "post_date": "07/30/2022 13:21:23",
      "content": "<p>Many people, myself included, are doing Kalman Smoothing, which effectively runs a Kalman filter forward, and then  backward over the results, combining the results appropriately.   This means that future data is used for positioning, and thus is not a real time method.</p>",
      "votes": null,
      "replies": [
        {
          "id": 1877471,
          "author_name": "leonshangguan",
          "author_url": "",
          "post_date": "07/30/2022 17:05:12",
          "content": "<p>Got it and thanks for the comment. I am wondering if it is possible to do real-time localization since I think that's more like our daily routine. <a href=\"https://www.kaggle.com/mohammedkhider\" target=\"_blank\">@mohammedkhider</a> Will you hold the same type of competition for next year? Do you think we can do real-time localization down to the decimeter? Thanks!</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1876887": "May I ask if your solutions are real-time localization or need post-processing? One of my recent projects requires real-time precise localization, but I followed some public open-source solutions and find most solutions require whole route information in this competition (i.e., access to all data after route completion) and then do filtering. I appreciate it if you can help.",
    "1877267": "Many people, myself included, are doing Kalman Smoothing, which effectively runs a Kalman filter forward, and then  backward over the results, combining the results appropriately.   This means that future data is used for positioning, and thus is not a real time method.",
    "1877471": "Got it and thanks for the comment. I am wondering if it is possible to do real-time localization since I think that's more like our daily routine. @mohammedkhider Will you hold the same type of competition for next year? Do you think we can do real-time localization down to the decimeter? Thanks!"
  },
  "source": "meta"
}