{
  "id": 343615,
  "title": "Download HuBMAP unlabeled data",
  "url": "/competitions/hubmap-organ-segmentation/discussion/343615",
  "author_name": "",
  "post_date": "2022-08-12T01:38:26.839213400Z",
  "votes": 16,
  "comment_count": 2,
  "views": 0,
  "content": "<p>I have downloaded and tested an image and cropped it to a size of 3000x3000. Because the images here are too big, I can't download them all at once, here is just a download sample, I hope it can help you. Currently I'm downloading at a rate of every 40 due to output storage limitations. If there is a better way to download all of them, we can discuss here.</p>\n<p>Here is my download notebook:<br>\n<a href=\"https://www.kaggle.com/code/yingpengchen/hubmap-data-download\" target=\"_blank\">https://www.kaggle.com/code/yingpengchen/hubmap-data-download</a> </p>\n<p>Please check version 2.</p>",
  "messages": [
    {
      "id": "1895138",
      "postDate": "08/12/2022 01:38:26",
      "content": "<p>I have downloaded and tested an image and cropped it to a size of 3000x3000. Because the images here are too big, I can't download them all at once, here is just a download sample, I hope it can help you. Currently I'm downloading at a rate of every 40 due to output storage limitations. If there is a better way to download all of them, we can discuss here.</p>\n<p>Here is my download notebook:<br>\n<a href=\"https://www.kaggle.com/code/yingpengchen/hubmap-data-download\" target=\"_blank\">https://www.kaggle.com/code/yingpengchen/hubmap-data-download</a> </p>\n<p>Please check version 2.</p>",
      "rawMarkdown": "I have downloaded and tested an image and cropped it to a size of 3000x3000. Because the images here are too big, I can't download them all at once, here is just a download sample, I hope it can help you. Currently I'm downloading at a rate of every 40 due to output storage limitations. If there is a better way to download all of them, we can discuss here.\n\nHere is my download notebook:\n[https://www.kaggle.com/code/yingpengchen/hubmap-data-download](https://www.kaggle.com/code/yingpengchen/hubmap-data-download) \n\nPlease check version 2.",
      "votes": null
    },
    {
      "id": "1895431",
      "postDate": "08/12/2022 06:13:28",
      "content": "<p>Thanks for sharing. I think downloading all of them isn't necessary. I can actually crop between 3 and 5 regions from each whole slide image manually. It could be more than 5 for prostate images since they can be cropped in lower resolution.  </p>",
      "rawMarkdown": "Thanks for sharing. I think downloading all of them isn't necessary. I can actually crop between 3 and 5 regions from each whole slide image manually. It could be more than 5 for prostate images since they can be cropped in lower resolution.",
      "votes": null
    },
    {
      "id": "1896177",
      "postDate": "08/12/2022 16:25:00",
      "content": "<p>Yes, because there are so many white spaces in hubmap dataset, we don't actually need to download them. So I tried to crop them and then decide if they were empty before downloading them.</p>",
      "rawMarkdown": "Yes, because there are so many white spaces in hubmap dataset, we don't actually need to download them. So I tried to crop them and then decide if they were empty before downloading them.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1895431,
      "author_name": "gunesevitan",
      "author_url": "",
      "post_date": "08/12/2022 06:13:28",
      "content": "<p>Thanks for sharing. I think downloading all of them isn't necessary. I can actually crop between 3 and 5 regions from each whole slide image manually. It could be more than 5 for prostate images since they can be cropped in lower resolution.  </p>",
      "votes": null,
      "replies": [
        {
          "id": 1896177,
          "author_name": "yingpengchen",
          "author_url": "",
          "post_date": "08/12/2022 16:25:00",
          "content": "<p>Yes, because there are so many white spaces in hubmap dataset, we don't actually need to download them. So I tried to crop them and then decide if they were empty before downloading them.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1895138": "I have downloaded and tested an image and cropped it to a size of 3000x3000. Because the images here are too big, I can't download them all at once, here is just a download sample, I hope it can help you. Currently I'm downloading at a rate of every 40 due to output storage limitations. If there is a better way to download all of them, we can discuss here.\n\nHere is my download notebook:\n[https://www.kaggle.com/code/yingpengchen/hubmap-data-download](https://www.kaggle.com/code/yingpengchen/hubmap-data-download) \n\nPlease check version 2.",
    "1895431": "Thanks for sharing. I think downloading all of them isn't necessary. I can actually crop between 3 and 5 regions from each whole slide image manually. It could be more than 5 for prostate images since they can be cropped in lower resolution.",
    "1896177": "Yes, because there are so many white spaces in hubmap dataset, we don't actually need to download them. So I tried to crop them and then decide if they were empty before downloading them."
  },
  "source": "meta"
}