{
  "id": 98636,
  "title": "Image Preprocessing",
  "url": "/competitions/aptos2019-blindness-detection/discussion/98636",
  "author_name": "",
  "post_date": "2019-07-05T09:52:51.412002900Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>I had a pretty stupid doubt. Given the number of errors in the data, is it advisable to preprocess our set of images(like denoise them,etc) or would it perform poorly with the private dataset</p>",
  "messages": [
    {
      "id": "568678",
      "postDate": "07/05/2019 09:52:51",
      "content": "<p>I had a pretty stupid doubt. Given the number of errors in the data, is it advisable to preprocess our set of images(like denoise them,etc) or would it perform poorly with the private dataset</p>",
      "rawMarkdown": "I had a pretty stupid doubt. Given the number of errors in the data, is it advisable to preprocess our set of images(like denoise them,etc) or would it perform poorly with the private dataset",
      "votes": null
    },
    {
      "id": "568728",
      "postDate": "07/05/2019 10:36:56",
      "content": "<p>I believe it is a good idea. Otherwise without denoising the images, the model performance will be worse.</p>",
      "rawMarkdown": "I believe it is a good idea. Otherwise without denoising the images, the model performance will be worse.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 568728,
      "author_name": "souravbhattacharya10",
      "author_url": "",
      "post_date": "07/05/2019 10:36:56",
      "content": "<p>I believe it is a good idea. Otherwise without denoising the images, the model performance will be worse.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "568678": "I had a pretty stupid doubt. Given the number of errors in the data, is it advisable to preprocess our set of images(like denoise them,etc) or would it perform poorly with the private dataset",
    "568728": "I believe it is a good idea. Otherwise without denoising the images, the model performance will be worse."
  },
  "source": "meta"
}