{
  "id": 226137,
  "title": "56.4 AP！Faster and stronger CenterNet2 is coming!",
  "url": "/competitions/vinbigdata-chest-xray-abnormalities-detection/discussion/226137",
  "author_name": "Hanson0910",
  "post_date": "2021-03-15T10:55:35.732000",
  "votes": 12,
  "comment_count": 1,
  "views": 0,
  "content": "<p>It can reach 56.4 map on coco only in single scale test! <br>\ncode link:<a href=\"url\" target=\"_blank\">https://github.com/xingyizhou/CenterNet2</a><br>\npaper link:<a href=\"url\" target=\"_blank\">https://arxiv.org/abs/2103.07461</a></p>",
  "messages": [
    {
      "id": 1238841,
      "postDate": "2021-03-15T10:55:35.733Z",
      "content": "<p>It can reach 56.4 map on coco only in single scale test! <br>\ncode link:<a href=\"url\" target=\"_blank\">https://github.com/xingyizhou/CenterNet2</a><br>\npaper link:<a href=\"url\" target=\"_blank\">https://arxiv.org/abs/2103.07461</a></p>",
      "rawMarkdown": "It can reach 56.4 map on coco only in single scale test! \ncode link:[https://github.com/xingyizhou/CenterNet2](url)\npaper link:[https://arxiv.org/abs/2103.07461](url)\n",
      "votes": 12
    },
    {
      "id": 1273114,
      "postDate": "2021-04-14T05:17:04.837Z",
      "content": "<p>Hi! I want to train a custom dataset using CenterNet2, but I found many problems in configuration from the detectron2. Has anyone come into the light related to custom training? </p>",
      "rawMarkdown": "Hi! I want to train a custom dataset using CenterNet2, but I found many problems in configuration from the detectron2. Has anyone come into the light related to custom training? "
    }
  ],
  "comments": [
    {
      "id": 1273114,
      "author_name": "Salman Chen",
      "author_url": "",
      "post_date": "2021-04-14T05:17:04.837000",
      "content": "<p>Hi! I want to train a custom dataset using CenterNet2, but I found many problems in configuration from the detectron2. Has anyone come into the light related to custom training? </p>",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1238841": "It can reach 56.4 map on coco only in single scale test! \ncode link:[https://github.com/xingyizhou/CenterNet2](url)\npaper link:[https://arxiv.org/abs/2103.07461](url)\n",
    "1273114": "Hi! I want to train a custom dataset using CenterNet2, but I found many problems in configuration from the detectron2. Has anyone come into the light related to custom training? "
  }
}