{
  "id": 240367,
  "title": "What type of method to use, R-CNN, YOLO, SSD or any other things?",
  "url": "/competitions/siim-covid19-detection/discussion/240367",
  "author_name": "",
  "post_date": "2021-05-19T13:41:59.127745100Z",
  "votes": 6,
  "comment_count": 1,
  "views": 0,
  "content": "<p>I am the beginner of object detection.</p>\n<p>I'm wondering what type of method to use.<br>\nThere are object detection methods like;</p>\n<ul>\n<li><strong>R-CNN</strong></li>\n<li><strong>YOLO</strong></li>\n<li><strong>SSD</strong></li>\n</ul>\n<p>and perhaps any other things🤔</p>\n<p>I think each has its advantages and disadvantages, but I don't know which one is suitable for this competition.<br>\nSo if you have any ideas, please share it🙇‍♂️</p>\n<p>And I have an additional question.<br>\nCan we use <strong>YOLO</strong> in this competition and some other of object detection?<br>\nI remember that there were discussions about its <strong>license</strong> in the past competitions.</p>\n<p>Thanks☺️</p>",
  "messages": [
    {
      "id": "1314997",
      "postDate": "05/19/2021 13:41:59",
      "content": "<p>I am the beginner of object detection.</p>\n<p>I'm wondering what type of method to use.<br>\nThere are object detection methods like;</p>\n<ul>\n<li><strong>R-CNN</strong></li>\n<li><strong>YOLO</strong></li>\n<li><strong>SSD</strong></li>\n</ul>\n<p>and perhaps any other things🤔</p>\n<p>I think each has its advantages and disadvantages, but I don't know which one is suitable for this competition.<br>\nSo if you have any ideas, please share it🙇‍♂️</p>\n<p>And I have an additional question.<br>\nCan we use <strong>YOLO</strong> in this competition and some other of object detection?<br>\nI remember that there were discussions about its <strong>license</strong> in the past competitions.</p>\n<p>Thanks☺️</p>",
      "rawMarkdown": "I am the beginner of object detection.\n\nI'm wondering what type of method to use.\nThere are object detection methods like;\n- **R-CNN**\n- **YOLO**\n- **SSD**\n\nand perhaps any other things🤔\n\nI think each has its advantages and disadvantages, but I don't know which one is suitable for this competition.\nSo if you have any ideas, please share it🙇‍♂️\n\nAnd I have an additional question.\nCan we use **YOLO** in this competition and some other of object detection?\nI remember that there were discussions about its **license** in the past competitions.\n\nThanks☺️",
      "votes": null
    },
    {
      "id": "1454189",
      "postDate": "08/06/2021 05:57:41",
      "content": "<p>Hey <a href=\"https://www.kaggle.com/yhirakawa\" target=\"_blank\">@yhirakawa</a>, I believe YOLO falls under the GNU Public License (at least<a href=\"https://github.com/ultralytics/yolov5/blob/master/LICENSE\" target=\"_blank\"> this one does</a>) which is permitted by the competition rules!</p>\n<p>I'd also like to learn more about the advantages of different object detection methods specifically for medical images.</p>\n<p>From what I understand, YOLO struggles with detecting small objects because it can only identify a single class per grid. This may not be the optimal approach since lung nodules in X-ray images are likely very small. Therefore, Faster R-CNN might be the best approach in terms of accuracy. Read more <a href=\"https://cv-tricks.com/object-detection/faster-r-cnn-yolo-ssd/\" target=\"_blank\">here</a>!</p>",
      "rawMarkdown": "Hey @yhirakawa, I believe YOLO falls under the GNU Public License (at least[ this one does](https://github.com/ultralytics/yolov5/blob/master/LICENSE)) which is permitted by the competition rules!\n\nI'd also like to learn more about the advantages of different object detection methods specifically for medical images.\n\nFrom what I understand, YOLO struggles with detecting small objects because it can only identify a single class per grid. This may not be the optimal approach since lung nodules in X-ray images are likely very small. Therefore, Faster R-CNN might be the best approach in terms of accuracy. Read more [here](https://cv-tricks.com/object-detection/faster-r-cnn-yolo-ssd/)!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1454189,
      "author_name": "ecthompson99",
      "author_url": "",
      "post_date": "08/06/2021 05:57:41",
      "content": "<p>Hey <a href=\"https://www.kaggle.com/yhirakawa\" target=\"_blank\">@yhirakawa</a>, I believe YOLO falls under the GNU Public License (at least<a href=\"https://github.com/ultralytics/yolov5/blob/master/LICENSE\" target=\"_blank\"> this one does</a>) which is permitted by the competition rules!</p>\n<p>I'd also like to learn more about the advantages of different object detection methods specifically for medical images.</p>\n<p>From what I understand, YOLO struggles with detecting small objects because it can only identify a single class per grid. This may not be the optimal approach since lung nodules in X-ray images are likely very small. Therefore, Faster R-CNN might be the best approach in terms of accuracy. Read more <a href=\"https://cv-tricks.com/object-detection/faster-r-cnn-yolo-ssd/\" target=\"_blank\">here</a>!</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1314997": "I am the beginner of object detection.\n\nI'm wondering what type of method to use.\nThere are object detection methods like;\n- **R-CNN**\n- **YOLO**\n- **SSD**\n\nand perhaps any other things🤔\n\nI think each has its advantages and disadvantages, but I don't know which one is suitable for this competition.\nSo if you have any ideas, please share it🙇‍♂️\n\nAnd I have an additional question.\nCan we use **YOLO** in this competition and some other of object detection?\nI remember that there were discussions about its **license** in the past competitions.\n\nThanks☺️",
    "1454189": "Hey @yhirakawa, I believe YOLO falls under the GNU Public License (at least[ this one does](https://github.com/ultralytics/yolov5/blob/master/LICENSE)) which is permitted by the competition rules!\n\nI'd also like to learn more about the advantages of different object detection methods specifically for medical images.\n\nFrom what I understand, YOLO struggles with detecting small objects because it can only identify a single class per grid. This may not be the optimal approach since lung nodules in X-ray images are likely very small. Therefore, Faster R-CNN might be the best approach in terms of accuracy. Read more [here](https://cv-tricks.com/object-detection/faster-r-cnn-yolo-ssd/)!"
  },
  "source": "meta"
}