{
  "id": 249517,
  "title": "Understanding Competition Submission File",
  "url": "/competitions/siim-covid19-detection/discussion/249517",
  "author_name": "",
  "post_date": "2021-06-28T18:27:20.372682200Z",
  "votes": 1,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hi everyone, </p>\n<p>For successful submissions, you have to predict opacity and bounding boxes along with labels (Please correct me if I am wrong). However, I am unable to wrap my head around it. Predicting labels makes sense but what is opacity? and why do we have to predict it and how can we predict it? I do not understand the overall nature of the submission file that we have to create.  What is its physical significance?</p>\n<p>I am very new to deep learning and this is my first project when it comes to using DL for medical image classification. So I am sorry in advance for the lame question. </p>\n<p>Help appreciated.<br>\nThanks Kanishk </p>",
  "messages": [
    {
      "id": "1368634",
      "postDate": "06/28/2021 18:27:20",
      "content": "<p>Hi everyone, </p>\n<p>For successful submissions, you have to predict opacity and bounding boxes along with labels (Please correct me if I am wrong). However, I am unable to wrap my head around it. Predicting labels makes sense but what is opacity? and why do we have to predict it and how can we predict it? I do not understand the overall nature of the submission file that we have to create.  What is its physical significance?</p>\n<p>I am very new to deep learning and this is my first project when it comes to using DL for medical image classification. So I am sorry in advance for the lame question. </p>\n<p>Help appreciated.<br>\nThanks Kanishk </p>",
      "rawMarkdown": "Hi everyone, \n\nFor successful submissions, you have to predict opacity and bounding boxes along with labels (Please correct me if I am wrong). However, I am unable to wrap my head around it. Predicting labels makes sense but what is opacity? and why do we have to predict it and how can we predict it? I do not understand the overall nature of the submission file that we have to create.  What is its physical significance?\n\nI am very new to deep learning and this is my first project when it comes to using DL for medical image classification. So I am sorry in advance for the lame question. \n\nHelp appreciated.\nThanks Kanishk",
      "votes": null
    },
    {
      "id": "1369472",
      "postDate": "06/29/2021 11:41:29",
      "content": "<p>Hi, I had the same difficulty in understanding what image-level and study-level meant. I found this <a href=\"https://www.kaggle.com/c/siim-covid19-detection/discussion/240878\" target=\"_blank\">notebook</a> helpful.</p>",
      "rawMarkdown": "Hi, I had the same difficulty in understanding what image-level and study-level meant. I found this [notebook](https://www.kaggle.com/c/siim-covid19-detection/discussion/240878) helpful.",
      "votes": null
    },
    {
      "id": "1369806",
      "postDate": "06/29/2021 15:45:42",
      "content": "<p>We have to predict two things. Study level and image level.</p>\n<ul>\n<li><p>Study level prediction is one of four classes .. Atypical, Indeterminate, Negative or Typical. They're mutually exclusive, but the instructions say there can be more than one class per study (this confuses me).</p></li>\n<li><p>Image level prediction is zero or more bounding boxes that represent opacities.</p></li>\n</ul>",
      "rawMarkdown": "We have to predict two things. Study level and image level.\n\n- Study level prediction is one of four classes .. Atypical, Indeterminate, Negative or Typical. They're mutually exclusive, but the instructions say there can be more than one class per study (this confuses me).\n\n- Image level prediction is zero or more bounding boxes that represent opacities.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1369472,
      "author_name": "thatgeeman",
      "author_url": "",
      "post_date": "06/29/2021 11:41:29",
      "content": "<p>Hi, I had the same difficulty in understanding what image-level and study-level meant. I found this <a href=\"https://www.kaggle.com/c/siim-covid19-detection/discussion/240878\" target=\"_blank\">notebook</a> helpful.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1369806,
      "author_name": "davidbroberts",
      "author_url": "",
      "post_date": "06/29/2021 15:45:42",
      "content": "<p>We have to predict two things. Study level and image level.</p>\n<ul>\n<li><p>Study level prediction is one of four classes .. Atypical, Indeterminate, Negative or Typical. They're mutually exclusive, but the instructions say there can be more than one class per study (this confuses me).</p></li>\n<li><p>Image level prediction is zero or more bounding boxes that represent opacities.</p></li>\n</ul>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1368634": "Hi everyone, \n\nFor successful submissions, you have to predict opacity and bounding boxes along with labels (Please correct me if I am wrong). However, I am unable to wrap my head around it. Predicting labels makes sense but what is opacity? and why do we have to predict it and how can we predict it? I do not understand the overall nature of the submission file that we have to create.  What is its physical significance?\n\nI am very new to deep learning and this is my first project when it comes to using DL for medical image classification. So I am sorry in advance for the lame question. \n\nHelp appreciated.\nThanks Kanishk",
    "1369472": "Hi, I had the same difficulty in understanding what image-level and study-level meant. I found this [notebook](https://www.kaggle.com/c/siim-covid19-detection/discussion/240878) helpful.",
    "1369806": "We have to predict two things. Study level and image level.\n\n- Study level prediction is one of four classes .. Atypical, Indeterminate, Negative or Typical. They're mutually exclusive, but the instructions say there can be more than one class per study (this confuses me).\n\n- Image level prediction is zero or more bounding boxes that represent opacities."
  },
  "source": "meta"
}