{
  "id": 127669,
  "title": "per image standardization rather than mean standardization?",
  "url": "/competitions/deepfake-detection-challenge/discussion/127669",
  "author_name": "hongy",
  "post_date": "2020-01-25T16:08:56.523000",
  "votes": 1,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Has anyone else experimented with per image standardization? </p>\n\n<p><a href=\"https://stackoverflow.com/questions/54155481/tensorflow-per-image-standardization-vs-mean-standardization-across-full-dataset\">https://stackoverflow.com/questions/54155481/tensorflow-per-image-standardization-vs-mean-standardization-across-full-dataset</a></p>\n\n<p>I think it works better, but I don't have much experience with images. </p>",
  "messages": [
    {
      "id": 729035,
      "postDate": "2020-01-25T16:08:56.523Z",
      "content": "<p>Has anyone else experimented with per image standardization? </p>\n\n<p><a href=\"https://stackoverflow.com/questions/54155481/tensorflow-per-image-standardization-vs-mean-standardization-across-full-dataset\">https://stackoverflow.com/questions/54155481/tensorflow-per-image-standardization-vs-mean-standardization-across-full-dataset</a></p>\n\n<p>I think it works better, but I don't have much experience with images. </p>",
      "rawMarkdown": "Has anyone else experimented with per image standardization? \n\nhttps://stackoverflow.com/questions/54155481/tensorflow-per-image-standardization-vs-mean-standardization-across-full-dataset\n\nI think it works better, but I don't have much experience with images. \n\n",
      "votes": 1
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "729035": "Has anyone else experimented with per image standardization? \n\nhttps://stackoverflow.com/questions/54155481/tensorflow-per-image-standardization-vs-mean-standardization-across-full-dataset\n\nI think it works better, but I don't have much experience with images. \n\n"
  }
}