{
  "id": 371414,
  "title": "Yolov7 and understanding the boxes coordinates",
  "url": "/competitions/nfl-player-contact-detection/discussion/371414",
  "author_name": "Adrian Muresan",
  "post_date": "2022-12-09T20:18:52.775000",
  "votes": 4,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Hi all, </p>\n<p>I took one Bundesliga shoot-out notebook made by Tito <a href=\"https://www.kaggle.com/its7171\" target=\"_blank\">@its7171</a> , modified it to our competition and went a bit further: I saved the boxes coordinates in text files, converted them to top, bottom, left and right coordinates, and them plotted them again on a random frame to make sure I converted them correctly and I understand how the coordinates are given from the yolov7 algorithm. You will need the extracted boxes coordinates if you want to work with the tracking data. <br>\nHere's the notebook:<br>\n<a href=\"url\" target=\"_blank\">https://www.kaggle.com/code/louisbunuel/yolov7-nfl-contact</a></p>",
  "messages": [
    {
      "id": 2060368,
      "postDate": "2022-12-09T20:18:52.777Z",
      "content": "<p>Hi all, </p>\n<p>I took one Bundesliga shoot-out notebook made by Tito <a href=\"https://www.kaggle.com/its7171\" target=\"_blank\">@its7171</a> , modified it to our competition and went a bit further: I saved the boxes coordinates in text files, converted them to top, bottom, left and right coordinates, and them plotted them again on a random frame to make sure I converted them correctly and I understand how the coordinates are given from the yolov7 algorithm. You will need the extracted boxes coordinates if you want to work with the tracking data. <br>\nHere's the notebook:<br>\n<a href=\"url\" target=\"_blank\">https://www.kaggle.com/code/louisbunuel/yolov7-nfl-contact</a></p>",
      "rawMarkdown": "Hi all, \n\nI took one Bundesliga shoot-out notebook made by Tito @its7171 , modified it to our competition and went a bit further: I saved the boxes coordinates in text files, converted them to top, bottom, left and right coordinates, and them plotted them again on a random frame to make sure I converted them correctly and I understand how the coordinates are given from the yolov7 algorithm. You will need the extracted boxes coordinates if you want to work with the tracking data. \nHere's the notebook:\n[https://www.kaggle.com/code/louisbunuel/yolov7-nfl-contact](url)\n",
      "votes": 4
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "2060368": "Hi all, \n\nI took one Bundesliga shoot-out notebook made by Tito @its7171 , modified it to our competition and went a bit further: I saved the boxes coordinates in text files, converted them to top, bottom, left and right coordinates, and them plotted them again on a random frame to make sure I converted them correctly and I understand how the coordinates are given from the yolov7 algorithm. You will need the extracted boxes coordinates if you want to work with the tracking data. \nHere's the notebook:\n[https://www.kaggle.com/code/louisbunuel/yolov7-nfl-contact](url)\n"
  }
}