{
  "id": 346256,
  "title": "Any hint for secret-sloth?",
  "url": "/competitions/ai-village-ctf/discussion/346256",
  "author_name": "",
  "post_date": "2022-08-18T16:09:15.733478600Z",
  "votes": 3,
  "comment_count": 4,
  "views": 0,
  "content": "<p>Have been watching the picture for several days :(</p>",
  "messages": [
    {
      "id": "1904951",
      "postDate": "08/18/2022 16:09:15",
      "content": "<p>Have been watching the picture for several days :(</p>",
      "rawMarkdown": "Have been watching the picture for several days :(",
      "votes": null
    },
    {
      "id": "1904952",
      "postDate": "08/18/2022 16:11:36",
      "content": "<p>There is some good discussion in the discord (found in the overview), but no real hints. People have solved it, keep trying! </p>",
      "rawMarkdown": "There is some good discussion in the discord (found in the overview), but no real hints. People have solved it, keep trying!",
      "votes": null
    },
    {
      "id": "1904953",
      "postDate": "08/18/2022 16:12:47",
      "content": "<p>I try the blue channel in the array because it was created in Matplotlib and the hat is blue in the picture. But i found nothing. :)</p>",
      "rawMarkdown": "I try the blue channel in the array because it was created in Matplotlib and the hat is blue in the picture. But i found nothing. :)",
      "votes": null
    },
    {
      "id": "1905555",
      "postDate": "08/19/2022 06:22:46",
      "content": "<p>Haven't solved it yet, but looking at the enlarged picture by eye, there are two short suspicious looking pixel lines on the sloths arm.<br>\nBut maybe this is random chance noise or intended deception. <br>\nI tried some simple steganography methods on the image as a whole, but had no success. </p>\n<p>And as a general line of thought: The harder challenges seem to rely on seeing some hidden flaw or mechanism and fixing/exploiting it. As steganography is the obvious choice, it might not be the solution at all. But maybe I'm trying to think to complicated/higher level now without having the understanding - and therefore getting it wrong.</p>",
      "rawMarkdown": "Haven't solved it yet, but looking at the enlarged picture by eye, there are two short suspicious looking pixel lines on the sloths arm.\nBut maybe this is random chance noise or intended deception. \nI tried some simple steganography methods on the image as a whole, but had no success. \n\nAnd as a general line of thought: The harder challenges seem to rely on seeing some hidden flaw or mechanism and fixing/exploiting it. As steganography is the obvious choice, it might not be the solution at all. But maybe I'm trying to think to complicated/higher level now without having the understanding - and therefore getting it wrong.",
      "votes": null
    },
    {
      "id": "1905811",
      "postDate": "08/19/2022 10:24:02",
      "content": "<p>I think the main difficulty here is because there are too many possible path to take, and we're not sure which one is the correct path. E.g.</p>\n<ol>\n<li>Performing byte level analysis on the file.</li>\n<li>Performing analysis on the color/RGB data.</li>\n<li>Trying ML/NN approach related with Steganography.</li>\n<li>Any other?</li>\n</ol>\n<p>I think #1 is less likely. The #3 is too overkill, albeit it's still possible (this is Kaggle after all). So far I only tried #2 but have found too many suspicious and confusing things.</p>",
      "rawMarkdown": "I think the main difficulty here is because there are too many possible path to take, and we're not sure which one is the correct path. E.g.\n1. Performing byte level analysis on the file.\n2. Performing analysis on the color/RGB data.\n3. Trying ML/NN approach related with Steganography.\n4. Any other?\n\nI think #1 is less likely. The #3 is too overkill, albeit it's still possible (this is Kaggle after all). So far I only tried #2 but have found too many suspicious and confusing things.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1904952,
      "author_name": "moohax",
      "author_url": "",
      "post_date": "08/18/2022 16:11:36",
      "content": "<p>There is some good discussion in the discord (found in the overview), but no real hints. People have solved it, keep trying! </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1904953,
      "author_name": "ambrusattila",
      "author_url": "",
      "post_date": "08/18/2022 16:12:47",
      "content": "<p>I try the blue channel in the array because it was created in Matplotlib and the hat is blue in the picture. But i found nothing. :)</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1905555,
      "author_name": "cepheidq",
      "author_url": "",
      "post_date": "08/19/2022 06:22:46",
      "content": "<p>Haven't solved it yet, but looking at the enlarged picture by eye, there are two short suspicious looking pixel lines on the sloths arm.<br>\nBut maybe this is random chance noise or intended deception. <br>\nI tried some simple steganography methods on the image as a whole, but had no success. </p>\n<p>And as a general line of thought: The harder challenges seem to rely on seeing some hidden flaw or mechanism and fixing/exploiting it. As steganography is the obvious choice, it might not be the solution at all. But maybe I'm trying to think to complicated/higher level now without having the understanding - and therefore getting it wrong.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1905811,
      "author_name": "thariqnugrohotomo",
      "author_url": "",
      "post_date": "08/19/2022 10:24:02",
      "content": "<p>I think the main difficulty here is because there are too many possible path to take, and we're not sure which one is the correct path. E.g.</p>\n<ol>\n<li>Performing byte level analysis on the file.</li>\n<li>Performing analysis on the color/RGB data.</li>\n<li>Trying ML/NN approach related with Steganography.</li>\n<li>Any other?</li>\n</ol>\n<p>I think #1 is less likely. The #3 is too overkill, albeit it's still possible (this is Kaggle after all). So far I only tried #2 but have found too many suspicious and confusing things.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1904951": "Have been watching the picture for several days :(",
    "1904952": "There is some good discussion in the discord (found in the overview), but no real hints. People have solved it, keep trying!",
    "1904953": "I try the blue channel in the array because it was created in Matplotlib and the hat is blue in the picture. But i found nothing. :)",
    "1905555": "Haven't solved it yet, but looking at the enlarged picture by eye, there are two short suspicious looking pixel lines on the sloths arm.\nBut maybe this is random chance noise or intended deception. \nI tried some simple steganography methods on the image as a whole, but had no success. \n\nAnd as a general line of thought: The harder challenges seem to rely on seeing some hidden flaw or mechanism and fixing/exploiting it. As steganography is the obvious choice, it might not be the solution at all. But maybe I'm trying to think to complicated/higher level now without having the understanding - and therefore getting it wrong.",
    "1905811": "I think the main difficulty here is because there are too many possible path to take, and we're not sure which one is the correct path. E.g.\n1. Performing byte level analysis on the file.\n2. Performing analysis on the color/RGB data.\n3. Trying ML/NN approach related with Steganography.\n4. Any other?\n\nI think #1 is less likely. The #3 is too overkill, albeit it's still possible (this is Kaggle after all). So far I only tried #2 but have found too many suspicious and confusing things."
  },
  "source": "meta"
}