{
  "id": 240545,
  "title": "Covid-Net pretrained models on 16000 images ",
  "url": "/competitions/siim-covid19-detection/discussion/240545",
  "author_name": "Antonio Rueda-Toicen",
  "post_date": "2021-05-20T09:32:36.071000",
  "votes": 8,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Source code and Tensorflow pretrained models for <a href=\"https://arxiv.org/abs/2105.00256\" target=\"_blank\">COVID-Net CXR-S</a> a covid severity-assessment model from a group at the University of Waterloo. </p>\n<p>Seems interesting for transfer learning, the authors report that it was trained on 16000 CXR images from 15000 patients. Their results were also validated by radiologists. </p>\n<p><a href=\"https://github.com/lindawangg/COVID-Net\" target=\"_blank\">https://github.com/lindawangg/COVID-Net</a></p>\n<p><a href=\"https://github.com/lindawangg/COVID-Net/blob/master/docs/models.md\" target=\"_blank\">https://github.com/lindawangg/COVID-Net/blob/master/docs/models.md</a></p>",
  "messages": [
    {
      "id": 1316068,
      "postDate": "2021-05-20T09:32:36.070Z",
      "content": "<p>Source code and Tensorflow pretrained models for <a href=\"https://arxiv.org/abs/2105.00256\" target=\"_blank\">COVID-Net CXR-S</a> a covid severity-assessment model from a group at the University of Waterloo. </p>\n<p>Seems interesting for transfer learning, the authors report that it was trained on 16000 CXR images from 15000 patients. Their results were also validated by radiologists. </p>\n<p><a href=\"https://github.com/lindawangg/COVID-Net\" target=\"_blank\">https://github.com/lindawangg/COVID-Net</a></p>\n<p><a href=\"https://github.com/lindawangg/COVID-Net/blob/master/docs/models.md\" target=\"_blank\">https://github.com/lindawangg/COVID-Net/blob/master/docs/models.md</a></p>",
      "rawMarkdown": "Source code and Tensorflow pretrained models for [COVID-Net CXR-S](https://arxiv.org/abs/2105.00256) a covid severity-assessment model from a group at the University of Waterloo. \n\nSeems interesting for transfer learning, the authors report that it was trained on 16000 CXR images from 15000 patients. Their results were also validated by radiologists. \n\nhttps://github.com/lindawangg/COVID-Net\n\nhttps://github.com/lindawangg/COVID-Net/blob/master/docs/models.md\n\n\n",
      "votes": 8
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "1316068": "Source code and Tensorflow pretrained models for [COVID-Net CXR-S](https://arxiv.org/abs/2105.00256) a covid severity-assessment model from a group at the University of Waterloo. \n\nSeems interesting for transfer learning, the authors report that it was trained on 16000 CXR images from 15000 patients. Their results were also validated by radiologists. \n\nhttps://github.com/lindawangg/COVID-Net\n\nhttps://github.com/lindawangg/COVID-Net/blob/master/docs/models.md\n\n\n"
  }
}