{
  "id": 30342,
  "title": "Annotations quality",
  "url": "/competitions/intel-mobileodt-cervical-cancer-screening/discussion/30342",
  "author_name": "",
  "post_date": "2017-03-18T21:37:27.054982200Z",
  "votes": null,
  "comment_count": 5,
  "views": 0,
  "content": "<p>Of course, we should consider labels from train set as ground truth, but I am just curious about who performed manual annotations. Because some images confuse me a lot (espessialy, when I clearly see the third type, but true type is 1.</p>",
  "messages": [
    {
      "id": "168968",
      "postDate": "03/18/2017 21:37:27",
      "content": "<p>Of course, we should consider labels from train set as ground truth, but I am just curious about who performed manual annotations. Because some images confuse me a lot (espessialy, when I clearly see the third type, but true type is 1.</p>",
      "rawMarkdown": "Of course, we should consider labels from train set as ground truth, but I am just curious about who performed manual annotations. Because some images confuse me a lot (espessialy, when I clearly see the third type, but true type is 1.",
      "votes": null
    },
    {
      "id": "168972",
      "postDate": "03/18/2017 21:47:46",
      "content": "<p>Can you post the example? Distinction between Type 3 and 1 is fairly straight forward.</p>",
      "rawMarkdown": "Can you post the example? Distinction between Type 3 and 1 is fairly straight forward.",
      "votes": null
    },
    {
      "id": "168983",
      "postDate": "03/18/2017 23:34:40",
      "content": "<p>Sure,\n109.jpg Type 1\n<img src=\"https://psv4.userapi.com/c810626/u24932420/docs/975681823b80/109.jpg\" alt=\"109.jpg\" title=\"\"></p>",
      "rawMarkdown": "Sure,\n109.jpg Type 1\n![109.jpg][1]\n\n\n  [1]: https://psv4.userapi.com/c810626/u24932420/docs/975681823b80/109.jpg",
      "votes": null
    },
    {
      "id": "168985",
      "postDate": "03/18/2017 23:41:43",
      "content": "<p>My key point is that there are some examples which seem to break types methodology, so maybe images need to be double checked.</p>",
      "rawMarkdown": "My key point is that there are some examples which seem to break types methodology, so maybe images need to be double checked.",
      "votes": null
    },
    {
      "id": "169062",
      "postDate": "03/19/2017 11:55:19",
      "content": "<p>It is type 1. You can see the unicellular epithelial tissue in the middle (the red one). Arguably  not the best representation of type 1.</p>",
      "rawMarkdown": "It is type 1. You can see the unicellular epithelial tissue in the middle (the red one). Arguably  not the best representation of type 1.",
      "votes": null
    },
    {
      "id": "169731",
      "postDate": "03/22/2017 10:42:53",
      "content": "",
      "rawMarkdown": "",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 168972,
      "author_name": "visoft",
      "author_url": "",
      "post_date": "03/18/2017 21:47:46",
      "content": "<p>Can you post the example? Distinction between Type 3 and 1 is fairly straight forward.</p>",
      "votes": null,
      "replies": [
        {
          "id": 168983,
          "author_name": "wolhow123",
          "author_url": "",
          "post_date": "03/18/2017 23:34:40",
          "content": "<p>Sure,\n109.jpg Type 1\n<img src=\"https://psv4.userapi.com/c810626/u24932420/docs/975681823b80/109.jpg\" alt=\"109.jpg\" title=\"\"></p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 169062,
          "author_name": "visoft",
          "author_url": "",
          "post_date": "03/19/2017 11:55:19",
          "content": "<p>It is type 1. You can see the unicellular epithelial tissue in the middle (the red one). Arguably  not the best representation of type 1.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 169731,
          "author_name": "neboduus",
          "author_url": "",
          "post_date": "03/22/2017 10:42:53",
          "content": "",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 168985,
      "author_name": "wolhow123",
      "author_url": "",
      "post_date": "03/18/2017 23:41:43",
      "content": "<p>My key point is that there are some examples which seem to break types methodology, so maybe images need to be double checked.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "168968": "Of course, we should consider labels from train set as ground truth, but I am just curious about who performed manual annotations. Because some images confuse me a lot (espessialy, when I clearly see the third type, but true type is 1.",
    "168972": "Can you post the example? Distinction between Type 3 and 1 is fairly straight forward.",
    "168983": "Sure,\n109.jpg Type 1\n![109.jpg][1]\n\n\n  [1]: https://psv4.userapi.com/c810626/u24932420/docs/975681823b80/109.jpg",
    "168985": "My key point is that there are some examples which seem to break types methodology, so maybe images need to be double checked.",
    "169062": "It is type 1. You can see the unicellular epithelial tissue in the middle (the red one). Arguably  not the best representation of type 1.",
    "169731": ""
  },
  "source": "meta"
}