{
  "id": 151862,
  "title": "EncodedPixels",
  "url": "/competitions/imaterialist-fashion-2020-fgvc7/discussion/151862",
  "author_name": "",
  "post_date": "2020-05-17T13:05:16.686708100Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Can anyone please explain what does the EncodedPixels column stores in layman terms. Thank you. Much Appreciated...</p>",
  "messages": [
    {
      "id": "851222",
      "postDate": "05/17/2020 13:05:16",
      "content": "<p>Can anyone please explain what does the EncodedPixels column stores in layman terms. Thank you. Much Appreciated...</p>",
      "rawMarkdown": "Can anyone please explain what does the EncodedPixels column stores in layman terms. Thank you. Much Appreciated...",
      "votes": null
    },
    {
      "id": "851759",
      "postDate": "05/17/2020 22:23:19",
      "content": "<p>This is the simplest explanation I found for EncodedPixels column in both training and test csv. Hope it helps!</p>\n\n<p><a href=\"https://www.kaggle.com/c/imaterialist-fashion-2020-fgvc7/overview/evaluation\">https://www.kaggle.com/c/imaterialist-fashion-2020-fgvc7/overview/evaluation</a>\nEncodedPixels\nIn order to reduce the submission file size, our metric uses run-length encoding on the pixel values. Instead of submitting an exhaustive list of indices for your segmentation, you will submit pairs of values that contain a start position and a run length. E.g. '1 3' implies starting at pixel 1 and running a total of 3 pixels (1,2,3).</p>\n\n<p>The competition format requires a space-delimited list of pairs. For example, '1 3 10 5' implies pixels 1,2,3,10,11,12,13,14 are to be included in the mask. The pixels are one-indexed and numbered from top to bottom, then left to right: 1 is pixel (1,1), 2 is pixel (2,1), etc.</p>\n\n<p>The metric checks that the pairs are sorted, positive, and the decoded pixel values are not duplicated. It also checks that no two predicted masks for the same class and the same image are overlapping.</p>",
      "rawMarkdown": "This is the simplest explanation I found for EncodedPixels column in both training and test csv. Hope it helps!\n\nhttps://www.kaggle.com/c/imaterialist-fashion-2020-fgvc7/overview/evaluation\nEncodedPixels\nIn order to reduce the submission file size, our metric uses run-length encoding on the pixel values. Instead of submitting an exhaustive list of indices for your segmentation, you will submit pairs of values that contain a start position and a run length. E.g. '1 3' implies starting at pixel 1 and running a total of 3 pixels (1,2,3).\n\nThe competition format requires a space-delimited list of pairs. For example, '1 3 10 5' implies pixels 1,2,3,10,11,12,13,14 are to be included in the mask. The pixels are one-indexed and numbered from top to bottom, then left to right: 1 is pixel (1,1), 2 is pixel (2,1), etc.\n\nThe metric checks that the pairs are sorted, positive, and the decoded pixel values are not duplicated. It also checks that no two predicted masks for the same class and the same image are overlapping.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 851759,
      "author_name": "mihirmavalankar",
      "author_url": "",
      "post_date": "05/17/2020 22:23:19",
      "content": "<p>This is the simplest explanation I found for EncodedPixels column in both training and test csv. Hope it helps!</p>\n\n<p><a href=\"https://www.kaggle.com/c/imaterialist-fashion-2020-fgvc7/overview/evaluation\">https://www.kaggle.com/c/imaterialist-fashion-2020-fgvc7/overview/evaluation</a>\nEncodedPixels\nIn order to reduce the submission file size, our metric uses run-length encoding on the pixel values. Instead of submitting an exhaustive list of indices for your segmentation, you will submit pairs of values that contain a start position and a run length. E.g. '1 3' implies starting at pixel 1 and running a total of 3 pixels (1,2,3).</p>\n\n<p>The competition format requires a space-delimited list of pairs. For example, '1 3 10 5' implies pixels 1,2,3,10,11,12,13,14 are to be included in the mask. The pixels are one-indexed and numbered from top to bottom, then left to right: 1 is pixel (1,1), 2 is pixel (2,1), etc.</p>\n\n<p>The metric checks that the pairs are sorted, positive, and the decoded pixel values are not duplicated. It also checks that no two predicted masks for the same class and the same image are overlapping.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "851222": "Can anyone please explain what does the EncodedPixels column stores in layman terms. Thank you. Much Appreciated...",
    "851759": "This is the simplest explanation I found for EncodedPixels column in both training and test csv. Hope it helps!\n\nhttps://www.kaggle.com/c/imaterialist-fashion-2020-fgvc7/overview/evaluation\nEncodedPixels\nIn order to reduce the submission file size, our metric uses run-length encoding on the pixel values. Instead of submitting an exhaustive list of indices for your segmentation, you will submit pairs of values that contain a start position and a run length. E.g. '1 3' implies starting at pixel 1 and running a total of 3 pixels (1,2,3).\n\nThe competition format requires a space-delimited list of pairs. For example, '1 3 10 5' implies pixels 1,2,3,10,11,12,13,14 are to be included in the mask. The pixels are one-indexed and numbered from top to bottom, then left to right: 1 is pixel (1,1), 2 is pixel (2,1), etc.\n\nThe metric checks that the pairs are sorted, positive, and the decoded pixel values are not duplicated. It also checks that no two predicted masks for the same class and the same image are overlapping."
  },
  "source": "meta"
}