{
  "id": 239952,
  "title": "Opacity label?",
  "url": "/competitions/siim-covid19-detection/discussion/239952",
  "author_name": "",
  "post_date": "2021-05-18T06:01:40.825097900Z",
  "votes": 8,
  "comment_count": 5,
  "views": 0,
  "content": "<p>What exactly is 'opacity' supposed to mean? </p>\n<blockquote>\n  <p>Further, for each test study, you should make a determination within the following labels:</p>\n  <p>Negative for Pneumonia' 'Typical Appearance' 'Indeterminate Appearance' 'Atypical Appearance</p>\n  <p>To make a prediction of one of the above labels, create a prediction string similar to the \"none\" class above: e.g. atypical 1 0 0 1 1</p>\n</blockquote>\n<p>As per the details on the Data page, predictions need to have label as one of negative, typical, indeterminate,atypical.</p>\n<p>The evaluation page says - </p>\n<p>The submission file should contain a header and have the following format:</p>\n<blockquote>\n  <p>Id,PredictionString<br>\n  2b95d54e4be65_study,negative 1 0 0 1 1<br>\n  2b95d54e4be66_study,typical 1 0 0 1 1<br>\n  2b95d54e4be66_study,indeterminate 1 0 0 1 1 atypical 1 0 0 1 1<br>\n  2b95d54e4be68_image,none 1 0 0 1 1<br>\n  2b95d54e4be69_image,opacity 0.5 100 100 200 200 opacity 0.7 10 10 20 20<br>\n  etc.</p>\n</blockquote>\n<p>Where opacity is also included as a label.</p>\n<p>And all the labels in training data contain opacity as the label. Does anyone know about this?<br>\nOr did I miss out on something important?</p>",
  "messages": [
    {
      "id": "1312575",
      "postDate": "05/18/2021 06:01:40",
      "content": "<p>What exactly is 'opacity' supposed to mean? </p>\n<blockquote>\n  <p>Further, for each test study, you should make a determination within the following labels:</p>\n  <p>Negative for Pneumonia' 'Typical Appearance' 'Indeterminate Appearance' 'Atypical Appearance</p>\n  <p>To make a prediction of one of the above labels, create a prediction string similar to the \"none\" class above: e.g. atypical 1 0 0 1 1</p>\n</blockquote>\n<p>As per the details on the Data page, predictions need to have label as one of negative, typical, indeterminate,atypical.</p>\n<p>The evaluation page says - </p>\n<p>The submission file should contain a header and have the following format:</p>\n<blockquote>\n  <p>Id,PredictionString<br>\n  2b95d54e4be65_study,negative 1 0 0 1 1<br>\n  2b95d54e4be66_study,typical 1 0 0 1 1<br>\n  2b95d54e4be66_study,indeterminate 1 0 0 1 1 atypical 1 0 0 1 1<br>\n  2b95d54e4be68_image,none 1 0 0 1 1<br>\n  2b95d54e4be69_image,opacity 0.5 100 100 200 200 opacity 0.7 10 10 20 20<br>\n  etc.</p>\n</blockquote>\n<p>Where opacity is also included as a label.</p>\n<p>And all the labels in training data contain opacity as the label. Does anyone know about this?<br>\nOr did I miss out on something important?</p>",
      "rawMarkdown": "What exactly is 'opacity' supposed to mean? \n\n>  Further, for each test study, you should make a determination within the following labels:\n> \n> Negative for Pneumonia' 'Typical Appearance' 'Indeterminate Appearance' 'Atypical Appearance\n> \n> To make a prediction of one of the above labels, create a prediction string similar to the \"none\" class above: e.g. atypical 1 0 0 1 1\n\nAs per the details on the Data page, predictions need to have label as one of negative, typical, indeterminate,atypical.\n\nThe evaluation page says - \n\nThe submission file should contain a header and have the following format:\n\n> Id,PredictionString\n2b95d54e4be65_study,negative 1 0 0 1 1\n2b95d54e4be66_study,typical 1 0 0 1 1\n2b95d54e4be66_study,indeterminate 1 0 0 1 1 atypical 1 0 0 1 1\n2b95d54e4be68_image,none 1 0 0 1 1\n2b95d54e4be69_image,opacity 0.5 100 100 200 200 opacity 0.7 10 10 20 20\netc.\n\nWhere opacity is also included as a label.\n\nAnd all the labels in training data contain opacity as the label. Does anyone know about this?\nOr did I miss out on something important?",
      "votes": null
    },
    {
      "id": "1313938",
      "postDate": "05/18/2021 20:08:25",
      "content": "<p>Hi! There are image- and study-level labels; \"opacity\" and \"none\" are image-level labels that apply to \"_image\" samples. The \"opacity\" label should be used with a bounding box that indicates an opacity on the image, \"none\" if there are none for a given image. This is noted under \"Image-level labels\" on the Evaluation page.</p>\n<p>The other four labels - \"negative\", \"atypical\", \"typical\", and \"indeterminate\" - are study-level labels and apply to \"_study\" samples.</p>\n<p>Check out the Data and Evaluation page for more details about the labels!</p>",
      "rawMarkdown": "Hi! There are image- and study-level labels; \"opacity\" and \"none\" are image-level labels that apply to \"_image\" samples. The \"opacity\" label should be used with a bounding box that indicates an opacity on the image, \"none\" if there are none for a given image. This is noted under \"Image-level labels\" on the Evaluation page.\n\nThe other four labels - \"negative\", \"atypical\", \"typical\", and \"indeterminate\" - are study-level labels and apply to \"_study\" samples.\n\nCheck out the Data and Evaluation page for more details about the labels!",
      "votes": null
    },
    {
      "id": "1313942",
      "postDate": "05/18/2021 20:12:08",
      "content": "<p>It was a bit confusing initially, but your explanation makes it clear now. Thank you!</p>",
      "rawMarkdown": "It was a bit confusing initially, but your explanation makes it clear now. Thank you!",
      "votes": null
    },
    {
      "id": "1385125",
      "postDate": "07/12/2021 13:25:35",
      "content": "<p>so can I confirm </p>\n<p>negative in study level == none in image level. <br>\nopacity in image level will be further classified into atypical, typical and indeterminate at study level? </p>",
      "rawMarkdown": "so can I confirm \n\nnegative in study level == none in image level. \nopacity in image level will be further classified into atypical, typical and indeterminate at study level?",
      "votes": null
    },
    {
      "id": "1385162",
      "postDate": "07/12/2021 13:52:46",
      "content": "<p><a href=\"https://www.kaggle.com/h053473666/siim-cov19-efnb7-yolov5-infer#1354525\" target=\"_blank\">https://www.kaggle.com/h053473666/siim-cov19-efnb7-yolov5-infer#1354525</a></p>\n<blockquote>\n  <p>none and Negative are not exactly the same</p>\n</blockquote>",
      "rawMarkdown": "https://www.kaggle.com/h053473666/siim-cov19-efnb7-yolov5-infer#1354525\n\n> none and Negative are not exactly the same",
      "votes": null
    },
    {
      "id": "1385220",
      "postDate": "07/12/2021 14:25:50",
      "content": "<p>Thanks. Now you reminded me this.</p>",
      "rawMarkdown": "Thanks. Now you reminded me this.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1313938,
      "author_name": "philculliton",
      "author_url": "",
      "post_date": "05/18/2021 20:08:25",
      "content": "<p>Hi! There are image- and study-level labels; \"opacity\" and \"none\" are image-level labels that apply to \"_image\" samples. The \"opacity\" label should be used with a bounding box that indicates an opacity on the image, \"none\" if there are none for a given image. This is noted under \"Image-level labels\" on the Evaluation page.</p>\n<p>The other four labels - \"negative\", \"atypical\", \"typical\", and \"indeterminate\" - are study-level labels and apply to \"_study\" samples.</p>\n<p>Check out the Data and Evaluation page for more details about the labels!</p>",
      "votes": null,
      "replies": [
        {
          "id": 1313942,
          "author_name": "p4rallax",
          "author_url": "",
          "post_date": "05/18/2021 20:12:08",
          "content": "<p>It was a bit confusing initially, but your explanation makes it clear now. Thank you!</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1385125,
          "author_name": "hawkeat",
          "author_url": "",
          "post_date": "07/12/2021 13:25:35",
          "content": "<p>so can I confirm </p>\n<p>negative in study level == none in image level. <br>\nopacity in image level will be further classified into atypical, typical and indeterminate at study level? </p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1385162,
          "author_name": "mknzfr",
          "author_url": "",
          "post_date": "07/12/2021 13:52:46",
          "content": "<p><a href=\"https://www.kaggle.com/h053473666/siim-cov19-efnb7-yolov5-infer#1354525\" target=\"_blank\">https://www.kaggle.com/h053473666/siim-cov19-efnb7-yolov5-infer#1354525</a></p>\n<blockquote>\n  <p>none and Negative are not exactly the same</p>\n</blockquote>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1385220,
          "author_name": "hawkeat",
          "author_url": "",
          "post_date": "07/12/2021 14:25:50",
          "content": "<p>Thanks. Now you reminded me this.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1312575": "What exactly is 'opacity' supposed to mean? \n\n>  Further, for each test study, you should make a determination within the following labels:\n> \n> Negative for Pneumonia' 'Typical Appearance' 'Indeterminate Appearance' 'Atypical Appearance\n> \n> To make a prediction of one of the above labels, create a prediction string similar to the \"none\" class above: e.g. atypical 1 0 0 1 1\n\nAs per the details on the Data page, predictions need to have label as one of negative, typical, indeterminate,atypical.\n\nThe evaluation page says - \n\nThe submission file should contain a header and have the following format:\n\n> Id,PredictionString\n2b95d54e4be65_study,negative 1 0 0 1 1\n2b95d54e4be66_study,typical 1 0 0 1 1\n2b95d54e4be66_study,indeterminate 1 0 0 1 1 atypical 1 0 0 1 1\n2b95d54e4be68_image,none 1 0 0 1 1\n2b95d54e4be69_image,opacity 0.5 100 100 200 200 opacity 0.7 10 10 20 20\netc.\n\nWhere opacity is also included as a label.\n\nAnd all the labels in training data contain opacity as the label. Does anyone know about this?\nOr did I miss out on something important?",
    "1313938": "Hi! There are image- and study-level labels; \"opacity\" and \"none\" are image-level labels that apply to \"_image\" samples. The \"opacity\" label should be used with a bounding box that indicates an opacity on the image, \"none\" if there are none for a given image. This is noted under \"Image-level labels\" on the Evaluation page.\n\nThe other four labels - \"negative\", \"atypical\", \"typical\", and \"indeterminate\" - are study-level labels and apply to \"_study\" samples.\n\nCheck out the Data and Evaluation page for more details about the labels!",
    "1313942": "It was a bit confusing initially, but your explanation makes it clear now. Thank you!",
    "1385125": "so can I confirm \n\nnegative in study level == none in image level. \nopacity in image level will be further classified into atypical, typical and indeterminate at study level?",
    "1385162": "https://www.kaggle.com/h053473666/siim-cov19-efnb7-yolov5-infer#1354525\n\n> none and Negative are not exactly the same",
    "1385220": "Thanks. Now you reminded me this."
  },
  "source": "meta"
}