{
  "id": 298559,
  "title": "test dataset size?",
  "url": "/competitions/tensorflow-great-barrier-reef/discussion/298559",
  "author_name": "",
  "post_date": "2022-01-03T15:11:25.214032500Z",
  "votes": null,
  "comment_count": 4,
  "views": 0,
  "content": "<p>submission sometime runs about half hours, sometimes about one hour.</p>\n<p>anyone else observes same situation?</p>",
  "messages": [
    {
      "id": "1637106",
      "postDate": "01/03/2022 15:11:25",
      "content": "<p>submission sometime runs about half hours, sometimes about one hour.</p>\n<p>anyone else observes same situation?</p>",
      "rawMarkdown": "submission sometime runs about half hours, sometimes about one hour.\n\n\nanyone else observes same situation?",
      "votes": null
    },
    {
      "id": "1637672",
      "postDate": "01/04/2022 05:41:10",
      "content": "<p>For me, it's always &gt; 30 mins</p>",
      "rawMarkdown": "For me, it's always > 30 mins",
      "votes": null
    },
    {
      "id": "1638064",
      "postDate": "01/04/2022 12:16:28",
      "content": "<p>I think It depends on how many submission you are running at the same time, I submit 4 submission and it takes longer</p>",
      "rawMarkdown": "I think It depends on how many submission you are running at the same time, I submit 4 submission and it takes longer",
      "votes": null
    },
    {
      "id": "1638254",
      "postDate": "01/04/2022 15:10:08",
      "content": "<p><a href=\"https://arxiv.org/pdf/2111.14311.pdf\" target=\"_blank\">https://arxiv.org/pdf/2111.14311.pdf</a><br>\nThe CSIRO Crown-of-Thorn Starfish Detection Dataset - Jiajun Liu</p>\n<p>\"collected at five different areas on a reef in the GBR\"<br>\n\" It contains more than 35k images, with hundreds of individual COTS visible.\"</p>\n<p>… there is a leak … now we know there are about hundreds of unique COTS objects (each one can appear in multiple frames)</p>",
      "rawMarkdown": "https://arxiv.org/pdf/2111.14311.pdf\nThe CSIRO Crown-of-Thorn Starfish Detection Dataset - Jiajun Liu\n\n\"collected at five different areas on a reef in the GBR\"\n\" It contains more than 35k images, with hundreds of individual COTS visible.\"\n\n... there is a leak ... now we know there are about hundreds of unique COTS objects (each one can appear in multiple frames)",
      "votes": null
    },
    {
      "id": "1639117",
      "postDate": "01/05/2022 11:48:28",
      "content": "<p>there is another leak</p>\n<p>\"The dataset then underwent an AI-assisted annotation and quality assurance process, in which expert annotators, with the help of pre-trained COTS detection models, identified all COTS in the<br>\nimages. Each COTS detection in an image was marked with a rectangular box using our annotation software.\"</p>\n<p>if we can model the output distribution \"pre-trained COTS detection models\" …. which usually is possible …</p>\n<p>\"The dataset was collected over a single day in October 2021 at a reef in the Swain Reefs region<br>\nof the GBR. There are variations in lighting, visibility, coral habitat, depth, distance from bottom and viewpoint.\" </p>\n<p>this is how to do augmentation</p>",
      "rawMarkdown": "there is another leak\n\n\"The dataset then underwent an AI-assisted annotation and quality assurance process, in which expert annotators, with the help of pre-trained COTS detection models, identified all COTS in the\nimages. Each COTS detection in an image was marked with a rectangular box using our annotation software.\"\n\nif we can model the output distribution \"pre-trained COTS detection models\" .... which usually is possible ...\n\n\n\"The dataset was collected over a single day in October 2021 at a reef in the Swain Reefs region\nof the GBR. There are variations in lighting, visibility, coral habitat, depth, distance from bottom and viewpoint.\" \n\nthis is how to do augmentation",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1637672,
      "author_name": "owenxing",
      "author_url": "",
      "post_date": "01/04/2022 05:41:10",
      "content": "<p>For me, it's always &gt; 30 mins</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1638064,
      "author_name": "ptran1203",
      "author_url": "",
      "post_date": "01/04/2022 12:16:28",
      "content": "<p>I think It depends on how many submission you are running at the same time, I submit 4 submission and it takes longer</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1638254,
      "author_name": "hengck23",
      "author_url": "",
      "post_date": "01/04/2022 15:10:08",
      "content": "<p><a href=\"https://arxiv.org/pdf/2111.14311.pdf\" target=\"_blank\">https://arxiv.org/pdf/2111.14311.pdf</a><br>\nThe CSIRO Crown-of-Thorn Starfish Detection Dataset - Jiajun Liu</p>\n<p>\"collected at five different areas on a reef in the GBR\"<br>\n\" It contains more than 35k images, with hundreds of individual COTS visible.\"</p>\n<p>… there is a leak … now we know there are about hundreds of unique COTS objects (each one can appear in multiple frames)</p>",
      "votes": null,
      "replies": [
        {
          "id": 1639117,
          "author_name": "hengck23",
          "author_url": "",
          "post_date": "01/05/2022 11:48:28",
          "content": "<p>there is another leak</p>\n<p>\"The dataset then underwent an AI-assisted annotation and quality assurance process, in which expert annotators, with the help of pre-trained COTS detection models, identified all COTS in the<br>\nimages. Each COTS detection in an image was marked with a rectangular box using our annotation software.\"</p>\n<p>if we can model the output distribution \"pre-trained COTS detection models\" …. which usually is possible …</p>\n<p>\"The dataset was collected over a single day in October 2021 at a reef in the Swain Reefs region<br>\nof the GBR. There are variations in lighting, visibility, coral habitat, depth, distance from bottom and viewpoint.\" </p>\n<p>this is how to do augmentation</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1637106": "submission sometime runs about half hours, sometimes about one hour.\n\n\nanyone else observes same situation?",
    "1637672": "For me, it's always > 30 mins",
    "1638064": "I think It depends on how many submission you are running at the same time, I submit 4 submission and it takes longer",
    "1638254": "https://arxiv.org/pdf/2111.14311.pdf\nThe CSIRO Crown-of-Thorn Starfish Detection Dataset - Jiajun Liu\n\n\"collected at five different areas on a reef in the GBR\"\n\" It contains more than 35k images, with hundreds of individual COTS visible.\"\n\n... there is a leak ... now we know there are about hundreds of unique COTS objects (each one can appear in multiple frames)",
    "1639117": "there is another leak\n\n\"The dataset then underwent an AI-assisted annotation and quality assurance process, in which expert annotators, with the help of pre-trained COTS detection models, identified all COTS in the\nimages. Each COTS detection in an image was marked with a rectangular box using our annotation software.\"\n\nif we can model the output distribution \"pre-trained COTS detection models\" .... which usually is possible ...\n\n\n\"The dataset was collected over a single day in October 2021 at a reef in the Swain Reefs region\nof the GBR. There are variations in lighting, visibility, coral habitat, depth, distance from bottom and viewpoint.\" \n\nthis is how to do augmentation"
  },
  "source": "meta"
}