{
  "id": 138837,
  "title": "Video Reader with GPU",
  "url": "/competitions/deepfake-detection-challenge/discussion/138837",
  "author_name": "",
  "post_date": "2020-03-26T14:23:01.359996600Z",
  "votes": 8,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Nvidia have created Video Processing Framework - good tool. I extended its with some useful function for python.</p>\n\n<p>original github:\n<a href=\"https://github.com/NVIDIA/VideoProcessingFramework\">https://github.com/NVIDIA/VideoProcessingFramework</a>\nextended version:\n<a href=\"https://github.com/leigh-plt/VideoProcessingFramework\">https://github.com/leigh-plt/VideoProcessingFramework</a>\nExample of usage:\n<a href=\"https://colab.research.google.com/drive/1hQGhHbonLNf0a5e6Ykr5S0hYk_mFwilO\">https://colab.research.google.com/drive/1hQGhHbonLNf0a5e6Ykr5S0hYk_mFwilO</a></p>\n\n<p>DLPack wrapper push tensor w/o copy operation to tf in eager mode or pytorch.\nOutput tensor located on gpu! You don't need copy it to device.</p>\n\n<p>Don't work with kaggle. Kaggle docker OS too old, some packages can't be upgraded safely.</p>",
  "messages": [
    {
      "id": "787105",
      "postDate": "03/26/2020 14:23:01",
      "content": "<p>Nvidia have created Video Processing Framework - good tool. I extended its with some useful function for python.</p>\n\n<p>original github:\n<a href=\"https://github.com/NVIDIA/VideoProcessingFramework\">https://github.com/NVIDIA/VideoProcessingFramework</a>\nextended version:\n<a href=\"https://github.com/leigh-plt/VideoProcessingFramework\">https://github.com/leigh-plt/VideoProcessingFramework</a>\nExample of usage:\n<a href=\"https://colab.research.google.com/drive/1hQGhHbonLNf0a5e6Ykr5S0hYk_mFwilO\">https://colab.research.google.com/drive/1hQGhHbonLNf0a5e6Ykr5S0hYk_mFwilO</a></p>\n\n<p>DLPack wrapper push tensor w/o copy operation to tf in eager mode or pytorch.\nOutput tensor located on gpu! You don't need copy it to device.</p>\n\n<p>Don't work with kaggle. Kaggle docker OS too old, some packages can't be upgraded safely.</p>",
      "rawMarkdown": "Nvidia have created Video Processing Framework - good tool. I extended its with some useful function for python.\n\noriginal github:\nhttps://github.com/NVIDIA/VideoProcessingFramework\nextended version:\nhttps://github.com/leigh-plt/VideoProcessingFramework\nExample of usage:\nhttps://colab.research.google.com/drive/1hQGhHbonLNf0a5e6Ykr5S0hYk_mFwilO\n\nDLPack wrapper push tensor w/o copy operation to tf in eager mode or pytorch.\nOutput tensor located on gpu! You don't need copy it to device.\n\nDon't work with kaggle. Kaggle docker OS too old, some packages can't be upgraded safely.",
      "votes": null
    },
    {
      "id": "787552",
      "postDate": "03/26/2020 21:57:03",
      "content": "<p>thanks for sharing. it's a shame it doesn't work with kaggle. </p>",
      "rawMarkdown": "thanks for sharing. it's a shame it doesn't work with kaggle.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 787552,
      "author_name": "yimacs",
      "author_url": "",
      "post_date": "03/26/2020 21:57:03",
      "content": "<p>thanks for sharing. it's a shame it doesn't work with kaggle. </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "787105": "Nvidia have created Video Processing Framework - good tool. I extended its with some useful function for python.\n\noriginal github:\nhttps://github.com/NVIDIA/VideoProcessingFramework\nextended version:\nhttps://github.com/leigh-plt/VideoProcessingFramework\nExample of usage:\nhttps://colab.research.google.com/drive/1hQGhHbonLNf0a5e6Ykr5S0hYk_mFwilO\n\nDLPack wrapper push tensor w/o copy operation to tf in eager mode or pytorch.\nOutput tensor located on gpu! You don't need copy it to device.\n\nDon't work with kaggle. Kaggle docker OS too old, some packages can't be upgraded safely.",
    "787552": "thanks for sharing. it's a shame it doesn't work with kaggle."
  },
  "source": "meta"
}