{
  "id": 405259,
  "title": "Loss functions for image segmentation",
  "url": "/competitions/vesuvius-challenge-ink-detection/discussion/405259",
  "author_name": "",
  "post_date": "2023-04-26T18:49:17.539270700Z",
  "votes": 43,
  "comment_count": 6,
  "views": 0,
  "content": "<p>Hi everyone! I found this site: <a href=\"https://github.com/JunMa11/SegLoss\" target=\"_blank\">https://github.com/JunMa11/SegLoss</a> where they summarize the segmentation loss types. It was useful for me, I hope it can help to others as well.</p>\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F4367831%2F717956da11a8953a81ec75a79495929c%2FLossOverview.jpg?generation=1609672056006545&amp;alt=media\" alt=\"\"></p>\n<p>Good luck for the competition!</p>",
  "messages": [
    {
      "id": "2236327",
      "postDate": "04/26/2023 18:49:17",
      "content": "<p>Hi everyone! I found this site: <a href=\"https://github.com/JunMa11/SegLoss\" target=\"_blank\">https://github.com/JunMa11/SegLoss</a> where they summarize the segmentation loss types. It was useful for me, I hope it can help to others as well.</p>\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F4367831%2F717956da11a8953a81ec75a79495929c%2FLossOverview.jpg?generation=1609672056006545&amp;alt=media\" alt=\"\"></p>\n<p>Good luck for the competition!</p>",
      "rawMarkdown": "Hi everyone! I found this site: https://github.com/JunMa11/SegLoss where they summarize the segmentation loss types. It was useful for me, I hope it can help to others as well.\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F4367831%2F717956da11a8953a81ec75a79495929c%2FLossOverview.jpg?generation=1609672056006545&alt=media)\n\nGood luck for the competition!",
      "votes": null
    },
    {
      "id": "2239175",
      "postDate": "04/29/2023 08:26:29",
      "content": "<p>Thank you for sharing! This is very useful! I will definitely use it!</p>",
      "rawMarkdown": "Thank you for sharing! This is very useful! I will definitely use it!",
      "votes": null
    },
    {
      "id": "2239505",
      "postDate": "04/29/2023 15:24:08",
      "content": "<p><a href=\"https://www.kaggle.com/bessenyeiszilrd\" target=\"_blank\">@bessenyeiszilrd</a> thank you for sharing this! Indeed it is very interesting and useful. I did use before only a few types of Cross entropy loss and IoU/Jaccard.😀 </p>",
      "rawMarkdown": "bessenyeiszilrd thank you for sharing this! Indeed it is very interesting and useful. I did use before only a few types of Cross entropy loss and IoU/Jaccard.😀",
      "votes": null
    },
    {
      "id": "2250619",
      "postDate": "05/08/2023 17:21:29",
      "content": "<p>Turned out to be quite useful. Thank you for posting this!</p>",
      "rawMarkdown": "Turned out to be quite useful. Thank you for posting this!",
      "votes": null
    },
    {
      "id": "2265949",
      "postDate": "05/19/2023 15:59:50",
      "content": "<p>Indeed very useful. Thanks for sharing!</p>",
      "rawMarkdown": "Indeed very useful. Thanks for sharing!",
      "votes": null
    },
    {
      "id": "2267264",
      "postDate": "05/20/2023 18:19:12",
      "content": "<p>Thanks for sharing. This is useful for me too.</p>",
      "rawMarkdown": "Thanks for sharing. This is useful for me too.",
      "votes": null
    },
    {
      "id": "2296422",
      "postDate": "06/11/2023 20:13:58",
      "content": "<p>That's a great diagram, bookmarking it. Thanks!</p>",
      "rawMarkdown": "That's a great diagram, bookmarking it. Thanks!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2239175,
      "author_name": "ericka42",
      "author_url": "",
      "post_date": "04/29/2023 08:26:29",
      "content": "<p>Thank you for sharing! This is very useful! I will definitely use it!</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 2239505,
      "author_name": "ivanisaev",
      "author_url": "",
      "post_date": "04/29/2023 15:24:08",
      "content": "<p><a href=\"https://www.kaggle.com/bessenyeiszilrd\" target=\"_blank\">@bessenyeiszilrd</a> thank you for sharing this! Indeed it is very interesting and useful. I did use before only a few types of Cross entropy loss and IoU/Jaccard.😀 </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 2250619,
      "author_name": "gregoryeritsyan",
      "author_url": "",
      "post_date": "05/08/2023 17:21:29",
      "content": "<p>Turned out to be quite useful. Thank you for posting this!</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 2265949,
      "author_name": "brunoetc",
      "author_url": "",
      "post_date": "05/19/2023 15:59:50",
      "content": "<p>Indeed very useful. Thanks for sharing!</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 2267264,
      "author_name": "maryamnoroozi68",
      "author_url": "",
      "post_date": "05/20/2023 18:19:12",
      "content": "<p>Thanks for sharing. This is useful for me too.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 2296422,
      "author_name": "yassinealouini",
      "author_url": "",
      "post_date": "06/11/2023 20:13:58",
      "content": "<p>That's a great diagram, bookmarking it. Thanks!</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2236327": "Hi everyone! I found this site: https://github.com/JunMa11/SegLoss where they summarize the segmentation loss types. It was useful for me, I hope it can help to others as well.\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F4367831%2F717956da11a8953a81ec75a79495929c%2FLossOverview.jpg?generation=1609672056006545&alt=media)\n\nGood luck for the competition!",
    "2239175": "Thank you for sharing! This is very useful! I will definitely use it!",
    "2239505": "bessenyeiszilrd thank you for sharing this! Indeed it is very interesting and useful. I did use before only a few types of Cross entropy loss and IoU/Jaccard.😀",
    "2250619": "Turned out to be quite useful. Thank you for posting this!",
    "2265949": "Indeed very useful. Thanks for sharing!",
    "2267264": "Thanks for sharing. This is useful for me too.",
    "2296422": "That's a great diagram, bookmarking it. Thanks!"
  },
  "source": "meta"
}