{
  "id": 153010,
  "title": "Splitting data based on Quality factor and Stego Scheme",
  "url": "/competitions/alaska2-image-steganalysis/discussion/153010",
  "author_name": "",
  "post_date": "2020-05-22T16:54:41.056198800Z",
  "votes": 8,
  "comment_count": 3,
  "views": 0,
  "content": "<p>Following the data-split strategy from <a href=\"http://www.ws.binghamton.edu/fridrich/Research/ALASKA-preprint1.pdf\">Alaska-I winners</a>,</p>\n\n<blockquote>\n  <p>The training set (TRN), validation set (VAL), and test set (TST) contained respectively 42,500, 3,500, and 3,500 cover images (around 500 cover images were not used because they were corrupted or failed the processing pipeline). The TRN, VAL, and TST sets were created for each quality factor and each stego scheme in TILEdouble, TILEbase, and ARBITRARYbase</p>\n</blockquote>\n\n<p>This <a href=\"https://www.kaggle.com/bibek777/datasplit-based-on-quality-factor-and-stego-scheme?scriptVersionId=34569586\">kernel</a> shows how you can split the dataset based on quality factor and Stego scheme.</p>",
  "messages": [
    {
      "id": "857508",
      "postDate": "05/22/2020 16:54:41",
      "content": "<p>Following the data-split strategy from <a href=\"http://www.ws.binghamton.edu/fridrich/Research/ALASKA-preprint1.pdf\">Alaska-I winners</a>,</p>\n\n<blockquote>\n  <p>The training set (TRN), validation set (VAL), and test set (TST) contained respectively 42,500, 3,500, and 3,500 cover images (around 500 cover images were not used because they were corrupted or failed the processing pipeline). The TRN, VAL, and TST sets were created for each quality factor and each stego scheme in TILEdouble, TILEbase, and ARBITRARYbase</p>\n</blockquote>\n\n<p>This <a href=\"https://www.kaggle.com/bibek777/datasplit-based-on-quality-factor-and-stego-scheme?scriptVersionId=34569586\">kernel</a> shows how you can split the dataset based on quality factor and Stego scheme.</p>",
      "rawMarkdown": "Following the data-split strategy from [Alaska-I winners](http://www.ws.binghamton.edu/fridrich/Research/ALASKA-preprint1.pdf),\n&gt; The training set (TRN), validation set (VAL), and test set (TST) contained respectively 42,500, 3,500, and 3,500 cover images (around 500 cover images were not used because they were corrupted or failed the processing pipeline). The TRN, VAL, and TST sets were created for each quality factor and each stego scheme in TILEdouble, TILEbase, and ARBITRARYbase\n\nThis [kernel](https://www.kaggle.com/bibek777/datasplit-based-on-quality-factor-and-stego-scheme?scriptVersionId=34569586) shows how you can split the dataset based on quality factor and Stego scheme.",
      "votes": null
    },
    {
      "id": "857611",
      "postDate": "05/22/2020 18:46:32",
      "content": "<p>great, thanks for sharing.</p>",
      "rawMarkdown": "great, thanks for sharing.",
      "votes": null
    },
    {
      "id": "859451",
      "postDate": "05/24/2020 12:42:45",
      "content": "<p>good to see you back Bibek . Thanks for the sharing. The quality factorwise division helps ..</p>",
      "rawMarkdown": "good to see you back Bibek . Thanks for the sharing. The quality factorwise division helps ..",
      "votes": null
    },
    {
      "id": "867868",
      "postDate": "05/30/2020 17:33:17",
      "content": "<p>Thanks!</p>",
      "rawMarkdown": "Thanks!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 857611,
      "author_name": "rohitsingh9990",
      "author_url": "",
      "post_date": "05/22/2020 18:46:32",
      "content": "<p>great, thanks for sharing.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 859451,
      "author_name": "phoenix9032",
      "author_url": "",
      "post_date": "05/24/2020 12:42:45",
      "content": "<p>good to see you back Bibek . Thanks for the sharing. The quality factorwise division helps ..</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 867868,
      "author_name": "vincentpoont2",
      "author_url": "",
      "post_date": "05/30/2020 17:33:17",
      "content": "<p>Thanks!</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "857508": "Following the data-split strategy from [Alaska-I winners](http://www.ws.binghamton.edu/fridrich/Research/ALASKA-preprint1.pdf),\n&gt; The training set (TRN), validation set (VAL), and test set (TST) contained respectively 42,500, 3,500, and 3,500 cover images (around 500 cover images were not used because they were corrupted or failed the processing pipeline). The TRN, VAL, and TST sets were created for each quality factor and each stego scheme in TILEdouble, TILEbase, and ARBITRARYbase\n\nThis [kernel](https://www.kaggle.com/bibek777/datasplit-based-on-quality-factor-and-stego-scheme?scriptVersionId=34569586) shows how you can split the dataset based on quality factor and Stego scheme.",
    "857611": "great, thanks for sharing.",
    "859451": "good to see you back Bibek . Thanks for the sharing. The quality factorwise division helps ..",
    "867868": "Thanks!"
  },
  "source": "meta"
}