{
  "id": 539733,
  "title": "Let me know if any questions or problems!",
  "url": "/competitions/predict-geographic-context-from-landscape-photos/discussion/539733",
  "author_name": "Barry Hunter",
  "post_date": "2024-10-10T14:31:43.229000",
  "votes": 0,
  "comment_count": 1,
  "views": 0,
  "content": "<p>This is first time produced something like this, played with some 'toy' datasets for tranfer learning, but not worked with any large datasets, so may not have set this up quite right. </p>\n<p>Creating a AI model for predicting these 'Context' Tags from the image, is something we think is feasible, although not sure how accurate it will be. The labeled  data which is user-contributed and is already of variable quality, and picking test general Context tags for animage, is very subjective process. </p>\n<p>… hoping others can make sense of this!</p>",
  "messages": [
    {
      "id": 3013790,
      "postDate": "2024-10-10T14:31:43.230Z",
      "content": "<p>This is first time produced something like this, played with some 'toy' datasets for tranfer learning, but not worked with any large datasets, so may not have set this up quite right. </p>\n<p>Creating a AI model for predicting these 'Context' Tags from the image, is something we think is feasible, although not sure how accurate it will be. The labeled  data which is user-contributed and is already of variable quality, and picking test general Context tags for animage, is very subjective process. </p>\n<p>… hoping others can make sense of this!</p>",
      "rawMarkdown": "This is first time produced something like this, played with some 'toy' datasets for tranfer learning, but not worked with any large datasets, so may not have set this up quite right. \n\nCreating a AI model for predicting these 'Context' Tags from the image, is something we think is feasible, although not sure how accurate it will be. The labeled  data which is user-contributed and is already of variable quality, and picking test general Context tags for animage, is very subjective process. \n\n... hoping others can make sense of this!\n"
    },
    {
      "id": 3032731,
      "postDate": "2024-10-31T09:49:12.740Z",
      "rawMarkdown": "",
      "isDeleted": true
    }
  ],
  "comments": [
    {
      "id": 3032731,
      "author_name": "",
      "author_url": "",
      "post_date": "2024-10-31T09:49:12.740000",
      "content": "",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3013790": "This is first time produced something like this, played with some 'toy' datasets for tranfer learning, but not worked with any large datasets, so may not have set this up quite right. \n\nCreating a AI model for predicting these 'Context' Tags from the image, is something we think is feasible, although not sure how accurate it will be. The labeled  data which is user-contributed and is already of variable quality, and picking test general Context tags for animage, is very subjective process. \n\n... hoping others can make sense of this!\n",
    "3032731": ""
  }
}