{
  "id": 238199,
  "title": "35th solution, seg&yolo",
  "url": "/competitions/hubmap-kidney-segmentation/writeups/babys-hacking-35th-solution-seg-yolo",
  "author_name": "",
  "post_date": "2021-05-11T13:47:23.724144800Z",
  "votes": 5,
  "comment_count": 2,
  "views": 0,
  "content": "<p>lb0.933, pb0.946<br>\ninference code：<a href=\"https://www.kaggle.com/kevin1742064161/seg-yolo-final-inference?scriptVersionId=62406590\" target=\"_blank\">https://www.kaggle.com/kevin1742064161/seg-yolo-final-inference?scriptVersionId=62406590</a><br>\nensemble of 4 model：unet+b2, unet+b4, unetplusplus+b2, manet+b4<br>\nwindow：1024<br>\noverlap：100<br>\nsize：512<br>\nYOLOv5 is used to detect cells. By comparing the area of the bonding box with the area of the segmentation, we can judge whether the segmentation result needs to be filled.<br>\nJust trust your CV!</p>",
  "messages": [
    {
      "id": "1302320",
      "postDate": "05/11/2021 13:47:23",
      "content": "<p>lb0.933, pb0.946<br>\ninference code：<a href=\"https://www.kaggle.com/kevin1742064161/seg-yolo-final-inference?scriptVersionId=62406590\" target=\"_blank\">https://www.kaggle.com/kevin1742064161/seg-yolo-final-inference?scriptVersionId=62406590</a><br>\nensemble of 4 model：unet+b2, unet+b4, unetplusplus+b2, manet+b4<br>\nwindow：1024<br>\noverlap：100<br>\nsize：512<br>\nYOLOv5 is used to detect cells. By comparing the area of the bonding box with the area of the segmentation, we can judge whether the segmentation result needs to be filled.<br>\nJust trust your CV!</p>",
      "rawMarkdown": "lb0.933, pb0.946\ninference code：https://www.kaggle.com/kevin1742064161/seg-yolo-final-inference?scriptVersionId=62406590\nensemble of 4 model：unet+b2, unet+b4, unetplusplus+b2, manet+b4\nwindow：1024\noverlap：100\nsize：512\nYOLOv5 is used to detect cells. By comparing the area of the bonding box with the area of the segmentation, we can judge whether the segmentation result needs to be filled.\nJust trust your CV!",
      "votes": null
    },
    {
      "id": "1303118",
      "postDate": "05/11/2021 23:17:34",
      "content": "<p><a href=\"https://www.kaggle.com/kevin1742064161\" target=\"_blank\">@kevin1742064161</a> Thanks for sharing the approach and Congratulations on winning medal</p>",
      "rawMarkdown": "kevin1742064161 Thanks for sharing the approach and Congratulations on winning medal",
      "votes": null
    },
    {
      "id": "1326552",
      "postDate": "05/28/2021 14:36:30",
      "content": "<p>I think it's a good idea to use the Object Detection model to find Cells. Did the detection &amp; segmentation models learn independently?</p>",
      "rawMarkdown": "I think it's a good idea to use the Object Detection model to find Cells. Did the detection & segmentation models learn independently?",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1303118,
      "author_name": "usharengaraju",
      "author_url": "",
      "post_date": "05/11/2021 23:17:34",
      "content": "<p><a href=\"https://www.kaggle.com/kevin1742064161\" target=\"_blank\">@kevin1742064161</a> Thanks for sharing the approach and Congratulations on winning medal</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1326552,
      "author_name": "chocozzz",
      "author_url": "",
      "post_date": "05/28/2021 14:36:30",
      "content": "<p>I think it's a good idea to use the Object Detection model to find Cells. Did the detection &amp; segmentation models learn independently?</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1302320": "lb0.933, pb0.946\ninference code：https://www.kaggle.com/kevin1742064161/seg-yolo-final-inference?scriptVersionId=62406590\nensemble of 4 model：unet+b2, unet+b4, unetplusplus+b2, manet+b4\nwindow：1024\noverlap：100\nsize：512\nYOLOv5 is used to detect cells. By comparing the area of the bonding box with the area of the segmentation, we can judge whether the segmentation result needs to be filled.\nJust trust your CV!",
    "1303118": "kevin1742064161 Thanks for sharing the approach and Congratulations on winning medal",
    "1326552": "I think it's a good idea to use the Object Detection model to find Cells. Did the detection & segmentation models learn independently?"
  },
  "source": "meta"
}