{
  "id": 228752,
  "title": "Resolution in this task.",
  "url": "/competitions/hubmap-kidney-segmentation/discussion/228752",
  "author_name": "Dylan Liu",
  "post_date": "2021-03-26T09:41:30.359000",
  "votes": 1,
  "comment_count": 0,
  "views": 0,
  "content": "<p>If we train the model on whole images, I can imagine that the larger the resolution is, the better the outcome will be.<br>\nBut in this kind of tasks we often collect training images with lower resolution from the whole images. So, how to properly set the resolution or image shape for this task? Or how does the resolution matter in this task?</p>",
  "messages": [
    {
      "id": 1253014,
      "postDate": "2021-03-26T09:41:30.360Z",
      "content": "<p>If we train the model on whole images, I can imagine that the larger the resolution is, the better the outcome will be.<br>\nBut in this kind of tasks we often collect training images with lower resolution from the whole images. So, how to properly set the resolution or image shape for this task? Or how does the resolution matter in this task?</p>",
      "rawMarkdown": "If we train the model on whole images, I can imagine that the larger the resolution is, the better the outcome will be.\nBut in this kind of tasks we often collect training images with lower resolution from the whole images. So, how to properly set the resolution or image shape for this task? Or how does the resolution matter in this task?",
      "votes": 1
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "1253014": "If we train the model on whole images, I can imagine that the larger the resolution is, the better the outcome will be.\nBut in this kind of tasks we often collect training images with lower resolution from the whole images. So, how to properly set the resolution or image shape for this task? Or how does the resolution matter in this task?"
  }
}