{
  "id": 20976,
  "title": "Manual labelling of train data",
  "url": "/competitions/state-farm-distracted-driver-detection/discussion/20976",
  "author_name": "",
  "post_date": "2016-05-15T15:28:40.123Z",
  "votes": null,
  "comment_count": 1,
  "views": 393,
  "content": "<p>With all these threads about &quot;external&quot; data I am wondering if it is &quot;legal&quot; to manually label the images in the train set?\nFor example if I make a &quot;smartphone detector&quot; I will manually crop them from a sample of train images, etc ..</p>",
  "messages": [
    {
      "id": "120111",
      "postDate": "05/15/2016 15:28:40",
      "content": "<p>With all these threads about &quot;external&quot; data I am wondering if it is &quot;legal&quot; to manually label the images in the train set?\nFor example if I make a &quot;smartphone detector&quot; I will manually crop them from a sample of train images, etc ..</p>",
      "rawMarkdown": "With all these threads about \"external\" data I am wondering if it is \"legal\" to manually label the images in the train set?\r\nFor example if I make a \"smartphone detector\" I will manually crop them from a sample of train images, etc ..",
      "votes": null
    },
    {
      "id": "120114",
      "postDate": "05/15/2016 16:08:56",
      "content": "<p>Yes this has been discussed and confirmed legit by competition admin here:  <a href=\"https://www.kaggle.com/c/state-farm-distracted-driver-detection/forums/t/19965/manual-labelling-of-training-data/114172#post114172\">https://www.kaggle.com/c/state-farm-distracted-driver-detection/forums/t/19965/manual-labelling-of-training-data/114172#post114172</a></p>",
      "rawMarkdown": "Yes this has been discussed and confirmed legit by competition admin here:  https://www.kaggle.com/c/state-farm-distracted-driver-detection/forums/t/19965/manual-labelling-of-training-data/114172#post114172",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 120114,
      "author_name": "slobo777",
      "author_url": "",
      "post_date": "05/15/2016 16:08:56",
      "content": "<p>Yes this has been discussed and confirmed legit by competition admin here:  <a href=\"https://www.kaggle.com/c/state-farm-distracted-driver-detection/forums/t/19965/manual-labelling-of-training-data/114172#post114172\">https://www.kaggle.com/c/state-farm-distracted-driver-detection/forums/t/19965/manual-labelling-of-training-data/114172#post114172</a></p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "120111": "With all these threads about \"external\" data I am wondering if it is \"legal\" to manually label the images in the train set?\r\nFor example if I make a \"smartphone detector\" I will manually crop them from a sample of train images, etc ..",
    "120114": "Yes this has been discussed and confirmed legit by competition admin here:  https://www.kaggle.com/c/state-farm-distracted-driver-detection/forums/t/19965/manual-labelling-of-training-data/114172#post114172"
  },
  "source": "meta"
}