{
  "id": 122018,
  "title": "Anyone created Image dataset by sampling?",
  "url": "/competitions/deepfake-detection-challenge/discussion/122018",
  "author_name": "Gaurav Gautam",
  "post_date": "2019-12-17T05:42:37.996000",
  "votes": 1,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Please share, If anyone created Image dataset from videos for whole data.\nAs we know that data size is huge. This dataset will be helpful and save time. So kindly share if created.</p>",
  "messages": [
    {
      "id": 696832,
      "postDate": "2019-12-17T05:42:37.997Z",
      "content": "<p>Please share, If anyone created Image dataset from videos for whole data.\nAs we know that data size is huge. This dataset will be helpful and save time. So kindly share if created.</p>",
      "rawMarkdown": "Please share, If anyone created Image dataset from videos for whole data.\nAs we know that data size is huge. This dataset will be helpful and save time. So kindly share if created.\n",
      "votes": 1
    },
    {
      "id": 715945,
      "postDate": "2020-01-11T02:40:24.140Z",
      "content": "<p><a href=\"/haodangmsu\">@haodangmsu</a> Thank you for this insight. Where in the challenge's specs do they state how videos are classified? E.g. does the existence of a single frame of FAKE in a video mean that a video is FAKE? Or, is the video classified by the classification of the majority of frames? (E.g. if the majority of frames are FAKE then the video is classified as fake)</p>\n\n<p>On the other hand, does this mean that every frame is REAL in all REAL videos?</p>",
      "rawMarkdown": "@haodangmsu Thank you for this insight. Where in the challenge's specs do they state how videos are classified? E.g. does the existence of a single frame of FAKE in a video mean that a video is FAKE? Or, is the video classified by the classification of the majority of frames? (E.g. if the majority of frames are FAKE then the video is classified as fake)\n\nOn the other hand, does this mean that every frame is REAL in all REAL videos?"
    },
    {
      "id": 713771,
      "postDate": "2020-01-08T16:28:57.523Z",
      "rawMarkdown": "",
      "votes": 1,
      "isDeleted": true
    }
  ],
  "comments": [
    {
      "id": 715945,
      "author_name": "Ayushi Sinha",
      "author_url": "",
      "post_date": "2020-01-11T02:40:24.140000",
      "content": "<p><a href=\"/haodangmsu\">@haodangmsu</a> Thank you for this insight. Where in the challenge's specs do they state how videos are classified? E.g. does the existence of a single frame of FAKE in a video mean that a video is FAKE? Or, is the video classified by the classification of the majority of frames? (E.g. if the majority of frames are FAKE then the video is classified as fake)</p>\n\n<p>On the other hand, does this mean that every frame is REAL in all REAL videos?</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 713771,
      "author_name": "",
      "author_url": "",
      "post_date": "2020-01-08T16:28:57.523000",
      "content": "",
      "votes": 1,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "696832": "Please share, If anyone created Image dataset from videos for whole data.\nAs we know that data size is huge. This dataset will be helpful and save time. So kindly share if created.\n",
    "715945": "@haodangmsu Thank you for this insight. Where in the challenge's specs do they state how videos are classified? E.g. does the existence of a single frame of FAKE in a video mean that a video is FAKE? Or, is the video classified by the classification of the majority of frames? (E.g. if the majority of frames are FAKE then the video is classified as fake)\n\nOn the other hand, does this mean that every frame is REAL in all REAL videos?",
    "713771": ""
  }
}