{
  "fetched_at": "2026-10-18T14:04:21Z",
  "source_url": "https://www.kaggle.com/competitions/deepfake-detection/data",
  "pages": [
    {
      "name": "data-description",
      "content": "### **Important** \nPlease submit your code once the competition ends. The submission deadline is 11:59 pm, 27th December 2021 (Monday). https://forms.gle/McFhWX4fYjENb73Y6\n\nThe dataset contains several mp4 videos which are both real and fake (AI-generated). The details of each folder can be found below. Your task is to predict videos present in the test folder as real or fake.\n\n## Mapping\n\n- Fake -> 0 \n- Real -> 1 \n\n## Files\n\n*   **deep-fake-dataset/Celeb-real** - the training dataset with label **1**\n*   **deep-fake-dataset/YouTube-real** - the training dataset with label **1**\n*   **deep-fake-dataset/Celeb-Youtube-fake** - training dataset with label **0**\n*   **deep-fake-dataset/test** - the test dataset\n*   **sample_submission.csv** - a sample submission file in the correct format\n\n## Columns\n\n*   `id` - video file name\n*   `label` - binary mapping to 0 or 1",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "Deepfake techniques, which present realistic AI-generated videos of people doing and saying fictional things, have the potential to have a significant impact on how people determine the legitimacy of information presented online. These content generation and modification technologies may affect the quality of public discourse and the safeguarding of human rights—especially given that deepfakes may be used maliciously as a source of misinformation, manipulation, harassment, and persuasion. Identifying manipulated media is a technically demanding and rapidly evolving challenge that requires collaborations across the entire tech industry and beyond.\n\nCOPS, IIT(BHU) presents a deepfake detection challenge which will help combat this menacing problem of deepfakes.\n\n### **Important** \n\nPlease submit your code once the competition ends. The submission deadline is 11:59 pm, 27th December 2021 (Monday). https://forms.gle/McFhWX4fYjENb73Y6",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Evaluation",
      "content": "The evaluation pages describe how submissions will be scored and how students should format their submissions. \n\nThe evaluation metric for this competition is [Mean F1-Score](https://en.wikipedia.org/wiki/F-score). The F1 score, commonly used in information retrieval, measures accuracy using the precision and recall of the statistics.\n\n![img](https://miro.medium.com/max/875/1*OhEnS-T54Cz0YSTl_c3Dwg.jpeg)\n![img](https://miro.medium.com/max/555/1*7J08ekAwupLBegeUI8muHA.png)\n\nPrecision is the ratio of true positives to all predicted positives. The recall is the ratio of true positives to all actual positives. \n\nThe F1 score is given by:\n\n![img](https://miro.medium.com/max/353/1*T6kVUKxG_Z4V5Fm1UXhEIw.png)\n\nThe F1 metric weights recall and precision equally, and a good retrieval algorithm will maximize both precision and recall simultaneously. Thus, moderately good performance on both will be favoured over extremely good performance on one and poor performance on the other.\n\n## Submission Format\nEach video present in the test folder must be mapped to a binary label (**0 for fake and 1 for real**). The submission will be in form of a csv file consisting of 2 columns which is shown below as a header.\n\nThe file should contain a header and have the following format:\n```\nid,label\n```\nThe **id** column contains the video name for example, 0.mp4 and the **label** column contains the video's binary mapping/classification 0 or 1.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "rules",
      "content": "- Maximum team size is 1\n- Please do not copy your code from other sources unless you completely understand them and can present them when asked to. Do not copy from other competitors, as you must submit your notebooks at the end. They need to reflect the exact result you submitted on the leaderboard and will be checked for plagiarism among other competitors. \n\n**Do not share your submission files with others**.\n\n**Do not forget to submit your code once the competition ends**.\n\n**It advised keeping the team name as the participant's name underscore roll number (ex: xyz_20000096). This would make it easier for us to keep a check and contact you**\n\n### **Important** \n\nPlease submit your code once the competition ends. The submission deadline is 11:59 pm, 27th December 2021 (Monday). https://forms.gle/McFhWX4fYjENb73Y6",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}