{
  "id": 101104,
  "title": "Handling Big Non-Square Image",
  "url": "/competitions/kuzushiji-recognition/discussion/101104",
  "author_name": "",
  "post_date": "2019-07-23T09:07:27.368313500Z",
  "votes": 2,
  "comment_count": 2,
  "views": 0,
  "content": "<p>How do you handle the data for this competition. What is the common practice for this kind of problem?</p>",
  "messages": [
    {
      "id": "582503",
      "postDate": "07/23/2019 09:07:27",
      "content": "<p>How do you handle the data for this competition. What is the common practice for this kind of problem?</p>",
      "rawMarkdown": "How do you handle the data for this competition. What is the common practice for this kind of problem?",
      "votes": null
    },
    {
      "id": "583758",
      "postDate": "07/25/2019 00:05:13",
      "content": "<p>Look up object detection algorithms, like RCNN, YOLO, etc.</p>",
      "rawMarkdown": "Look up object detection algorithms, like RCNN, YOLO, etc.",
      "votes": null
    },
    {
      "id": "584012",
      "postDate": "07/25/2019 10:03:49",
      "content": "<p>I have worked with those stuffs before but from my experience the object and image ratio are not as extreme as this, and image in my dataset is not as big(in terms of image size). Do you think it will work if I just resize the picture and jam in those algorithm?  My concern is only if I resize the image it will lost most of important details due to interpolation.</p>",
      "rawMarkdown": "I have worked with those stuffs before but from my experience the object and image ratio are not as extreme as this, and image in my dataset is not as big(in terms of image size). Do you think it will work if I just resize the picture and jam in those algorithm?  My concern is only if I resize the image it will lost most of important details due to interpolation.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 583758,
      "author_name": "sidhanthholalkere",
      "author_url": "",
      "post_date": "07/25/2019 00:05:13",
      "content": "<p>Look up object detection algorithms, like RCNN, YOLO, etc.</p>",
      "votes": null,
      "replies": [
        {
          "id": 584012,
          "author_name": "banzaibanzer",
          "author_url": "",
          "post_date": "07/25/2019 10:03:49",
          "content": "<p>I have worked with those stuffs before but from my experience the object and image ratio are not as extreme as this, and image in my dataset is not as big(in terms of image size). Do you think it will work if I just resize the picture and jam in those algorithm?  My concern is only if I resize the image it will lost most of important details due to interpolation.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "582503": "How do you handle the data for this competition. What is the common practice for this kind of problem?",
    "583758": "Look up object detection algorithms, like RCNN, YOLO, etc.",
    "584012": "I have worked with those stuffs before but from my experience the object and image ratio are not as extreme as this, and image in my dataset is not as big(in terms of image size). Do you think it will work if I just resize the picture and jam in those algorithm?  My concern is only if I resize the image it will lost most of important details due to interpolation."
  },
  "source": "meta"
}