{
  "id": 602424,
  "title": "Let's Discuss approaches ",
  "url": "/competitions/multi-class-object-detection-challenge/discussion/602424",
  "author_name": "Mohamed El-Nageeb",
  "post_date": "2025-08-27T07:28:23.993000",
  "votes": 0,
  "comment_count": 0,
  "views": 0,
  "content": "<ul>\n<li>Has anyone tried using two separate models per class?</li>\n</ul>\n<p>I experimented with this setup, thinking it would improve performance, but unfortunately it didn’t. Even when I included the top soup-cans model from the previous competition, results weren’t better.</p>\n<ul>\n<li>Did anyone try models other than YOLO ? </li>\n</ul>",
  "messages": [
    {
      "id": 3277030,
      "postDate": "2025-08-27T07:28:23.993Z",
      "content": "<ul>\n<li>Has anyone tried using two separate models per class?</li>\n</ul>\n<p>I experimented with this setup, thinking it would improve performance, but unfortunately it didn’t. Even when I included the top soup-cans model from the previous competition, results weren’t better.</p>\n<ul>\n<li>Did anyone try models other than YOLO ? </li>\n</ul>",
      "rawMarkdown": "- Has anyone tried using two separate models per class?\n\nI experimented with this setup, thinking it would improve performance, but unfortunately it didn’t. Even when I included the top soup-cans model from the previous competition, results weren’t better.\n\n- Did anyone try models other than YOLO ? "
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "3277030": "- Has anyone tried using two separate models per class?\n\nI experimented with this setup, thinking it would improve performance, but unfortunately it didn’t. Even when I included the top soup-cans model from the previous competition, results weren’t better.\n\n- Did anyone try models other than YOLO ? "
  }
}