{
  "id": 93020,
  "title": "Where are the bounding boxes?",
  "url": "/competitions/iwildcam-2019-fgvc6/discussion/93020",
  "author_name": "",
  "post_date": "2019-05-22T12:20:32.241650300Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>The data explanation mentions boxes for this dataset and the previous ones but I can't find it, anyone knows?</p>",
  "messages": [
    {
      "id": "535165",
      "postDate": "05/22/2019 12:20:32",
      "content": "<p>The data explanation mentions boxes for this dataset and the previous ones but I can't find it, anyone knows?</p>",
      "rawMarkdown": "The data explanation mentions boxes for this dataset and the previous ones but I can't find it, anyone knows?",
      "votes": null
    },
    {
      "id": "535341",
      "postDate": "05/22/2019 18:11:50",
      "content": "<p>The detection results are available on the <a href=\"https://github.com/visipedia/iwildcam_comp#camera-trap-animal-detection-model\">github page</a>.</p>",
      "rawMarkdown": "The detection results are available on the [github page](https://github.com/visipedia/iwildcam_comp#camera-trap-animal-detection-model).",
      "votes": null
    },
    {
      "id": "536424",
      "postDate": "05/24/2019 12:57:08",
      "content": "<p>Thank you for your reply. I saw and read the github page. I know that the iwild2019 is not complete therefore the participants have to make an extra effort trying to merge datasets into one (unless one just wants to train on different sub datasets). Not all datasets contain bounding boxes or any other annotation.</p>\n\n<p>The only annotations available are for</p>\n\n<p>iNat_Taxa_Map.json\n2019_CCT_Bboxes.json\n2019_CCT.json\n2019_IDFG_info.json \n2019_iNat_Idaho.json</p>\n\n<p>The iNat_Idaho.json has repeated images, so some categories have more images than in reality. Just to let you know. </p>",
      "rawMarkdown": "Thank you for your reply. I saw and read the github page. I know that the iwild2019 is not complete therefore the participants have to make an extra effort trying to merge datasets into one (unless one just wants to train on different sub datasets). Not all datasets contain bounding boxes or any other annotation.\n\nThe only annotations available are for\n\niNat_Taxa_Map.json\n2019_CCT_Bboxes.json\n2019_CCT.json\n2019_IDFG_info.json \n2019_iNat_Idaho.json\n\nThe iNat_Idaho.json has repeated images, so some categories have more images than in reality. Just to let you know.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 535341,
      "author_name": "sbeery",
      "author_url": "",
      "post_date": "05/22/2019 18:11:50",
      "content": "<p>The detection results are available on the <a href=\"https://github.com/visipedia/iwildcam_comp#camera-trap-animal-detection-model\">github page</a>.</p>",
      "votes": null,
      "replies": [
        {
          "id": 536424,
          "author_name": "lesolorzanov",
          "author_url": "",
          "post_date": "05/24/2019 12:57:08",
          "content": "<p>Thank you for your reply. I saw and read the github page. I know that the iwild2019 is not complete therefore the participants have to make an extra effort trying to merge datasets into one (unless one just wants to train on different sub datasets). Not all datasets contain bounding boxes or any other annotation.</p>\n\n<p>The only annotations available are for</p>\n\n<p>iNat_Taxa_Map.json\n2019_CCT_Bboxes.json\n2019_CCT.json\n2019_IDFG_info.json \n2019_iNat_Idaho.json</p>\n\n<p>The iNat_Idaho.json has repeated images, so some categories have more images than in reality. Just to let you know. </p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "535165": "The data explanation mentions boxes for this dataset and the previous ones but I can't find it, anyone knows?",
    "535341": "The detection results are available on the [github page](https://github.com/visipedia/iwildcam_comp#camera-trap-animal-detection-model).",
    "536424": "Thank you for your reply. I saw and read the github page. I know that the iwild2019 is not complete therefore the participants have to make an extra effort trying to merge datasets into one (unless one just wants to train on different sub datasets). Not all datasets contain bounding boxes or any other annotation.\n\nThe only annotations available are for\n\niNat_Taxa_Map.json\n2019_CCT_Bboxes.json\n2019_CCT.json\n2019_IDFG_info.json \n2019_iNat_Idaho.json\n\nThe iNat_Idaho.json has repeated images, so some categories have more images than in reality. Just to let you know."
  },
  "source": "meta"
}