{
  "id": 134921,
  "title": "A fun challenge for beginners: Predict if an image is GAN generated",
  "url": "/competitions/deepfake-detection-challenge/discussion/134921",
  "author_name": "xhlulu",
  "post_date": "2020-03-11T05:41:06.207000",
  "votes": 4,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Following up on my <a href=\"https://www.kaggle.com/c/deepfake-detection-challenge/discussion/129640\">previous post</a>, I decided to publicly release the <a href=\"https://www.kaggle.com/xhlulu/140k-real-and-fake-faces\">dataset</a>, and create <a href=\"https://www.kaggle.com/xhlulu/140k-real-and-fake-faces/tasks?taskId=530\">this task</a> for any beginner that want to get started with image classification, but that do not necessarily have the computation resources or technical expertise to deal with large amount of video data. </p>\n\n<p>The dataset was built on top of <a href=\"https://www.kaggle.com/c/deepfake-detection-challenge/discussion/122786\">the 70k real faces</a> dataset I released, as well as a subsample of the <a href=\"https://www.kaggle.com/c/deepfake-detection-challenge/discussion/121173\">1 million fake faces</a>  generated by Bojan. The dataset mostly just resize and rearrange both datasets all in one place, so that you can directly get started with your model.</p>\n\n<p>Please let me know your thoughts!</p>",
  "messages": [
    {
      "id": 768718,
      "postDate": "2020-03-11T05:41:06.207Z",
      "content": "<p>Following up on my <a href=\"https://www.kaggle.com/c/deepfake-detection-challenge/discussion/129640\">previous post</a>, I decided to publicly release the <a href=\"https://www.kaggle.com/xhlulu/140k-real-and-fake-faces\">dataset</a>, and create <a href=\"https://www.kaggle.com/xhlulu/140k-real-and-fake-faces/tasks?taskId=530\">this task</a> for any beginner that want to get started with image classification, but that do not necessarily have the computation resources or technical expertise to deal with large amount of video data. </p>\n\n<p>The dataset was built on top of <a href=\"https://www.kaggle.com/c/deepfake-detection-challenge/discussion/122786\">the 70k real faces</a> dataset I released, as well as a subsample of the <a href=\"https://www.kaggle.com/c/deepfake-detection-challenge/discussion/121173\">1 million fake faces</a>  generated by Bojan. The dataset mostly just resize and rearrange both datasets all in one place, so that you can directly get started with your model.</p>\n\n<p>Please let me know your thoughts!</p>",
      "rawMarkdown": "Following up on my [previous post](https://www.kaggle.com/c/deepfake-detection-challenge/discussion/129640), I decided to publicly release the [dataset](https://www.kaggle.com/xhlulu/140k-real-and-fake-faces), and create [this task](https://www.kaggle.com/xhlulu/140k-real-and-fake-faces/tasks?taskId=530) for any beginner that want to get started with image classification, but that do not necessarily have the computation resources or technical expertise to deal with large amount of video data. \n\nThe dataset was built on top of [the 70k real faces](https://www.kaggle.com/c/deepfake-detection-challenge/discussion/122786) dataset I released, as well as a subsample of the [1 million fake faces](https://www.kaggle.com/c/deepfake-detection-challenge/discussion/121173)  generated by Bojan. The dataset mostly just resize and rearrange both datasets all in one place, so that you can directly get started with your model.\n\nPlease let me know your thoughts!",
      "votes": 4
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "768718": "Following up on my [previous post](https://www.kaggle.com/c/deepfake-detection-challenge/discussion/129640), I decided to publicly release the [dataset](https://www.kaggle.com/xhlulu/140k-real-and-fake-faces), and create [this task](https://www.kaggle.com/xhlulu/140k-real-and-fake-faces/tasks?taskId=530) for any beginner that want to get started with image classification, but that do not necessarily have the computation resources or technical expertise to deal with large amount of video data. \n\nThe dataset was built on top of [the 70k real faces](https://www.kaggle.com/c/deepfake-detection-challenge/discussion/122786) dataset I released, as well as a subsample of the [1 million fake faces](https://www.kaggle.com/c/deepfake-detection-challenge/discussion/121173)  generated by Bojan. The dataset mostly just resize and rearrange both datasets all in one place, so that you can directly get started with your model.\n\nPlease let me know your thoughts!"
  }
}