{
  "id": 160346,
  "title": "Question about Site 3 + test.csv",
  "url": "/competitions/birdsong-recognition/discussion/160346",
  "author_name": "",
  "post_date": "2020-06-20T22:45:01.811129400Z",
  "votes": 5,
  "comment_count": 3,
  "views": 0,
  "content": "<p>Hello,</p>\n\n<p>I have some troubles understanding some things about site 3 and the structure of test_audio.</p>\n\n<p>First Question:\nWhen it is said  : \"Accordingly, site 3 has relatively few rows in the test set and needs lower time resolution predictions.\"\nWhat does \"needs lower time resolution predictions.\" mean ?</p>\n\n<p>Second Question:\nSince site 3 is labeled at the file level, is what is expected as prediction the list of bird names that appear on the whole 10 min file ? </p>\n\n<p>Third question: \nHow do we get the mp3 file path for the test set ? are there sub-folders for each site or is it just something like '../input/birdsong-recognition/test_audio/audio_id.mp3' and audio_id being something like 0a997dff022e3ad9744d4e7bbf923288</p>\n\n<p>Fourth question:\nIf a 5 second segment does not contain any bird sound, do we predict nocall or an empty string ?</p>\n\n<p>Thank you ! </p>",
  "messages": [
    {
      "id": "894921",
      "postDate": "06/20/2020 22:45:01",
      "content": "<p>Hello,</p>\n\n<p>I have some troubles understanding some things about site 3 and the structure of test_audio.</p>\n\n<p>First Question:\nWhen it is said  : \"Accordingly, site 3 has relatively few rows in the test set and needs lower time resolution predictions.\"\nWhat does \"needs lower time resolution predictions.\" mean ?</p>\n\n<p>Second Question:\nSince site 3 is labeled at the file level, is what is expected as prediction the list of bird names that appear on the whole 10 min file ? </p>\n\n<p>Third question: \nHow do we get the mp3 file path for the test set ? are there sub-folders for each site or is it just something like '../input/birdsong-recognition/test_audio/audio_id.mp3' and audio_id being something like 0a997dff022e3ad9744d4e7bbf923288</p>\n\n<p>Fourth question:\nIf a 5 second segment does not contain any bird sound, do we predict nocall or an empty string ?</p>\n\n<p>Thank you ! </p>",
      "rawMarkdown": "Hello,\n\nI have some troubles understanding some things about site 3 and the structure of test_audio.\n\nFirst Question:\nWhen it is said  : \"Accordingly, site 3 has relatively few rows in the test set and needs lower time resolution predictions.\"\nWhat does \"needs lower time resolution predictions.\" mean ?\n\nSecond Question:\nSince site 3 is labeled at the file level, is what is expected as prediction the list of bird names that appear on the whole 10 min file ? \n\nThird question: \nHow do we get the mp3 file path for the test set ? are there sub-folders for each site or is it just something like '../input/birdsong-recognition/test_audio/audio_id.mp3' and audio_id being something like 0a997dff022e3ad9744d4e7bbf923288\n\nFourth question:\nIf a 5 second segment does not contain any bird sound, do we predict nocall or an empty string ?\n\nThank you !",
      "votes": null
    },
    {
      "id": "894941",
      "postDate": "06/20/2020 23:28:09",
      "content": "<p>This notebook answers question 3 : <a href=\"https://www.kaggle.com/cwthompson/birdsong-making-a-prediction\">https://www.kaggle.com/cwthompson/birdsong-making-a-prediction</a>\nIt seems to be '../input/birdsong-recognition/testaudio/' + audio_id + '.mp3'\nThanks <a href=\"/cwthompson\">@cwthompson</a> </p>",
      "rawMarkdown": "This notebook answers question 3 : https://www.kaggle.com/cwthompson/birdsong-making-a-prediction\nIt seems to be '../input/birdsong-recognition/testaudio/' + audio_id + '.mp3'\nThanks @cwthompson",
      "votes": null
    },
    {
      "id": "894945",
      "postDate": "06/20/2020 23:52:19",
      "content": "<ol>\n<li>They only have clip level annotation for <code>site_3</code>, therefore what we need to do is clip level annotation.</li>\n<li>Yes</li>\n<li>Yes, as confirmed in <a href=\"https://www.kaggle.com/c/birdsong-recognition/discussion/158987\">https://www.kaggle.com/c/birdsong-recognition/discussion/158987</a> , <a href=\"https://www.kaggle.com/c/birdsong-recognition/discussion/159123\">https://www.kaggle.com/c/birdsong-recognition/discussion/159123</a></li>\n</ol>",
      "rawMarkdown": "1. They only have clip level annotation for `site_3`, therefore what we need to do is clip level annotation.\n2. Yes\n3. Yes, as confirmed in https://www.kaggle.com/c/birdsong-recognition/discussion/158987 , https://www.kaggle.com/c/birdsong-recognition/discussion/159123",
      "votes": null
    },
    {
      "id": "895686",
      "postDate": "06/21/2020 14:44:16",
      "content": "<p>4th - nocall.</p>",
      "rawMarkdown": "4th - nocall.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 894941,
      "author_name": "CVxTz",
      "author_url": "",
      "post_date": "06/20/2020 23:28:09",
      "content": "<p>This notebook answers question 3 : <a href=\"https://www.kaggle.com/cwthompson/birdsong-making-a-prediction\">https://www.kaggle.com/cwthompson/birdsong-making-a-prediction</a>\nIt seems to be '../input/birdsong-recognition/testaudio/' + audio_id + '.mp3'\nThanks <a href=\"/cwthompson\">@cwthompson</a> </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 894945,
      "author_name": "hidehisaarai1213",
      "author_url": "",
      "post_date": "06/20/2020 23:52:19",
      "content": "<ol>\n<li>They only have clip level annotation for <code>site_3</code>, therefore what we need to do is clip level annotation.</li>\n<li>Yes</li>\n<li>Yes, as confirmed in <a href=\"https://www.kaggle.com/c/birdsong-recognition/discussion/158987\">https://www.kaggle.com/c/birdsong-recognition/discussion/158987</a> , <a href=\"https://www.kaggle.com/c/birdsong-recognition/discussion/159123\">https://www.kaggle.com/c/birdsong-recognition/discussion/159123</a></li>\n</ol>",
      "votes": null,
      "replies": []
    },
    {
      "id": 895686,
      "author_name": "dhruvrnaik",
      "author_url": "",
      "post_date": "06/21/2020 14:44:16",
      "content": "<p>4th - nocall.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "894921": "Hello,\n\nI have some troubles understanding some things about site 3 and the structure of test_audio.\n\nFirst Question:\nWhen it is said  : \"Accordingly, site 3 has relatively few rows in the test set and needs lower time resolution predictions.\"\nWhat does \"needs lower time resolution predictions.\" mean ?\n\nSecond Question:\nSince site 3 is labeled at the file level, is what is expected as prediction the list of bird names that appear on the whole 10 min file ? \n\nThird question: \nHow do we get the mp3 file path for the test set ? are there sub-folders for each site or is it just something like '../input/birdsong-recognition/test_audio/audio_id.mp3' and audio_id being something like 0a997dff022e3ad9744d4e7bbf923288\n\nFourth question:\nIf a 5 second segment does not contain any bird sound, do we predict nocall or an empty string ?\n\nThank you !",
    "894941": "This notebook answers question 3 : https://www.kaggle.com/cwthompson/birdsong-making-a-prediction\nIt seems to be '../input/birdsong-recognition/testaudio/' + audio_id + '.mp3'\nThanks @cwthompson",
    "894945": "1. They only have clip level annotation for `site_3`, therefore what we need to do is clip level annotation.\n2. Yes\n3. Yes, as confirmed in https://www.kaggle.com/c/birdsong-recognition/discussion/158987 , https://www.kaggle.com/c/birdsong-recognition/discussion/159123",
    "895686": "4th - nocall."
  },
  "source": "meta"
}