{
  "id": 494027,
  "title": "Duration of hidden test/unlabelled soundscapes",
  "url": "/competitions/birdclef-2024/discussion/494027",
  "author_name": "",
  "post_date": "2024-04-15T18:20:38.911012400Z",
  "votes": 3,
  "comment_count": 1,
  "views": 0,
  "content": "<p>The soundscapes in the unlabelled directory seems to be of various durations. My question to the community (and the organisers) is whether we should expect similar variations in duration for hidden test set also. Although, the competition data descriptions clearly says that they are of 4min, I am wondering if that can be safely assumed.<br>\n2nd question is, in case the hidden test soundscapes are less or greater than 4 min duration, are we supposed to still predict for all 48 chunks</p>",
  "messages": [
    {
      "id": "2753953",
      "postDate": "04/15/2024 18:20:38",
      "content": "<p>The soundscapes in the unlabelled directory seems to be of various durations. My question to the community (and the organisers) is whether we should expect similar variations in duration for hidden test set also. Although, the competition data descriptions clearly says that they are of 4min, I am wondering if that can be safely assumed.<br>\n2nd question is, in case the hidden test soundscapes are less or greater than 4 min duration, are we supposed to still predict for all 48 chunks</p>",
      "rawMarkdown": "The soundscapes in the unlabelled directory seems to be of various durations. My question to the community (and the organisers) is whether we should expect similar variations in duration for hidden test set also. Although, the competition data descriptions clearly says that they are of 4min, I am wondering if that can be safely assumed.\n2nd question is, in case the hidden test soundscapes are less or greater than 4 min duration, are we supposed to still predict for all 48 chunks",
      "votes": null
    },
    {
      "id": "2754851",
      "postDate": "04/16/2024 08:30:26",
      "content": "<p>The duration in test set is 4 minutes for every soundscapes. Unlike what we have in unlabeled soundscapes.</p>",
      "rawMarkdown": "The duration in test set is 4 minutes for every soundscapes. Unlike what we have in unlabeled soundscapes.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2754851,
      "author_name": "cpmpml",
      "author_url": "",
      "post_date": "04/16/2024 08:30:26",
      "content": "<p>The duration in test set is 4 minutes for every soundscapes. Unlike what we have in unlabeled soundscapes.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2753953": "The soundscapes in the unlabelled directory seems to be of various durations. My question to the community (and the organisers) is whether we should expect similar variations in duration for hidden test set also. Although, the competition data descriptions clearly says that they are of 4min, I am wondering if that can be safely assumed.\n2nd question is, in case the hidden test soundscapes are less or greater than 4 min duration, are we supposed to still predict for all 48 chunks",
    "2754851": "The duration in test set is 4 minutes for every soundscapes. Unlike what we have in unlabeled soundscapes."
  },
  "source": "meta"
}