{
  "id": 126786,
  "title": "Submission - Using external models",
  "url": "/competitions/deepfake-detection-challenge/discussion/126786",
  "author_name": "Srinivasan.Vasu89",
  "post_date": "2020-01-20T08:06:10.566000",
  "votes": 0,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Hi, I am starting the project and new to the topic. as I was looking through the forums, I was figuring out, Face detection models is been used (like MTCNN n resnet) in public kernels. However, as per the coding rules, are we supposed to used already pretrained models that are already available in Kaggle Datasets or should I develop my own model for face-detection. It might sound so naive, but I don't want to mess with the rules. I would be happy if anyone could clarify on the topic.</p>\n\n<p>In that case, how can I include the external model to submission. A briefing would be good.</p>\n\n<p>PS: This is my first kaggle competition.</p>",
  "messages": [
    {
      "id": 723610,
      "postDate": "2020-01-20T08:06:10.567Z",
      "content": "<p>Hi, I am starting the project and new to the topic. as I was looking through the forums, I was figuring out, Face detection models is been used (like MTCNN n resnet) in public kernels. However, as per the coding rules, are we supposed to used already pretrained models that are already available in Kaggle Datasets or should I develop my own model for face-detection. It might sound so naive, but I don't want to mess with the rules. I would be happy if anyone could clarify on the topic.</p>\n\n<p>In that case, how can I include the external model to submission. A briefing would be good.</p>\n\n<p>PS: This is my first kaggle competition.</p>",
      "rawMarkdown": "Hi, I am starting the project and new to the topic. as I was looking through the forums, I was figuring out, Face detection models is been used (like MTCNN n resnet) in public kernels. However, as per the coding rules, are we supposed to used already pretrained models that are already available in Kaggle Datasets or should I develop my own model for face-detection. It might sound so naive, but I don't want to mess with the rules. I would be happy if anyone could clarify on the topic.\n\nIn that case, how can I include the external model to submission. A briefing would be good.\n\nPS: This is my first kaggle competition."
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "723610": "Hi, I am starting the project and new to the topic. as I was looking through the forums, I was figuring out, Face detection models is been used (like MTCNN n resnet) in public kernels. However, as per the coding rules, are we supposed to used already pretrained models that are already available in Kaggle Datasets or should I develop my own model for face-detection. It might sound so naive, but I don't want to mess with the rules. I would be happy if anyone could clarify on the topic.\n\nIn that case, how can I include the external model to submission. A briefing would be good.\n\nPS: This is my first kaggle competition."
  }
}