{
  "id": 253021,
  "title": "Clarification regarding one pixel bounding box in studies's predictions",
  "url": "/competitions/siim-covid19-detection/discussion/253021",
  "author_name": "Anmol Bhatia",
  "post_date": "2021-07-14T17:25:47.895000",
  "votes": 0,
  "comment_count": 5,
  "views": 0,
  "content": "<p>With respect to the predictions of only the studies in the test data , what does the one pixel bounding box of 0 0 1 1 in the sample submission represent ? Where is there a question of a box in image classification ? Bear with me if I'm sounding dumb , I'm new to this .</p>",
  "messages": [
    {
      "id": 1388406,
      "postDate": "2021-07-14T23:01:39.780Z",
      "content": "<p>The bounding box of 0 0 1 1 represents that there is no bounding box (no tumor) in the image. This is also an image classification and object detection competition so we are classifying the x-ray as well as finding the locations of tumors. </p>",
      "rawMarkdown": "The bounding box of 0 0 1 1 represents that there is no bounding box (no tumor) in the image. This is also an image classification and object detection competition so we are classifying the x-ray as well as finding the locations of tumors. ",
      "votes": 2,
      "replies": [
        {
          "id": 1388557,
          "postDate": "2021-07-15T04:28:01.403Z",
          "content": "<p>So , for all the studies the bounding box will be 0 0 1 1 right ? <br>\nPS : Thx for the reply !</p>",
          "rawMarkdown": "So , for all the studies the bounding box will be 0 0 1 1 right ? \nPS : Thx for the reply !"
        },
        {
          "id": 1389059,
          "postDate": "2021-07-15T12:16:08.010Z",
          "content": "<p>Yup, and for images, we would use the format of image name, then \"opacity\", probability, xmin, ymin, xmax, ymax for each detected bounding box or \"none\", 1 0 0 1 1. </p>",
          "rawMarkdown": "Yup, and for images, we would use the format of image name, then \"opacity\", probability, xmin, ymin, xmax, ymax for each detected bounding box or \"none\", 1 0 0 1 1. ",
          "votes": 1
        },
        {
          "id": 1389093,
          "postDate": "2021-07-15T13:02:06.393Z",
          "content": "<p>Got it , thx a lot !!</p>",
          "rawMarkdown": "Got it , thx a lot !!"
        }
      ]
    },
    {
      "id": 1388162,
      "postDate": "2021-07-14T17:25:47.897Z",
      "content": "<p>With respect to the predictions of only the studies in the test data , what does the one pixel bounding box of 0 0 1 1 in the sample submission represent ? Where is there a question of a box in image classification ? Bear with me if I'm sounding dumb , I'm new to this .</p>",
      "rawMarkdown": "With respect to the predictions of only the studies in the test data , what does the one pixel bounding box of 0 0 1 1 in the sample submission represent ? Where is there a question of a box in image classification ? Bear with me if I'm sounding dumb , I'm new to this ."
    },
    {
      "id": 1388556,
      "postDate": "2021-07-15T04:27:26.420Z",
      "rawMarkdown": "",
      "isDeleted": true
    }
  ],
  "comments": [
    {
      "id": 1388406,
      "author_name": "Ari",
      "author_url": "",
      "post_date": "2021-07-14T23:01:39.780000",
      "content": "<p>The bounding box of 0 0 1 1 represents that there is no bounding box (no tumor) in the image. This is also an image classification and object detection competition so we are classifying the x-ray as well as finding the locations of tumors. </p>",
      "votes": 2,
      "replies": [
        {
          "id": 1388557,
          "author_name": "Anmol Bhatia",
          "author_url": "",
          "post_date": "2021-07-15T04:28:01.403000",
          "content": "<p>So , for all the studies the bounding box will be 0 0 1 1 right ? <br>\nPS : Thx for the reply !</p>",
          "votes": 0,
          "replies": []
        },
        {
          "id": 1389059,
          "author_name": "Ari",
          "author_url": "",
          "post_date": "2021-07-15T12:16:08.010000",
          "content": "<p>Yup, and for images, we would use the format of image name, then \"opacity\", probability, xmin, ymin, xmax, ymax for each detected bounding box or \"none\", 1 0 0 1 1. </p>",
          "votes": 1,
          "replies": []
        },
        {
          "id": 1389093,
          "author_name": "Anmol Bhatia",
          "author_url": "",
          "post_date": "2021-07-15T13:02:06.393000",
          "content": "<p>Got it , thx a lot !!</p>",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 1388556,
      "author_name": "",
      "author_url": "",
      "post_date": "2021-07-15T04:27:26.420000",
      "content": "",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1388406": "The bounding box of 0 0 1 1 represents that there is no bounding box (no tumor) in the image. This is also an image classification and object detection competition so we are classifying the x-ray as well as finding the locations of tumors. ",
    "1388162": "With respect to the predictions of only the studies in the test data , what does the one pixel bounding box of 0 0 1 1 in the sample submission represent ? Where is there a question of a box in image classification ? Bear with me if I'm sounding dumb , I'm new to this .",
    "1388556": ""
  }
}