{
  "id": 491111,
  "title": "Train image filename cleanup",
  "url": "/competitions/image-matching-challenge-2024/discussion/491111",
  "author_name": "",
  "post_date": "2024-04-04T17:14:41.230382200Z",
  "votes": 7,
  "comment_count": 1,
  "views": 0,
  "content": "<p>As some of you had noticed, <a href=\"https://www.kaggle.com/competitions/image-matching-challenge-2024/discussion/487806#2733466\" target=\"_blank\">there were some inconsistencies</a> between some <code>image_name</code> entries in the <code>train_labels.csv</code> and the actual file names. I've posted a new version of the dataset that standardizes both the filenames and metadata to remove those inconsistencies.</p>\n<p>There are still some unlabeled image files in the train set; we're not planning on removing those.</p>\n<p>Apologies for the confusion and thank you all for your patience.</p>",
  "messages": [
    {
      "id": "2735342",
      "postDate": "04/04/2024 17:14:41",
      "content": "<p>As some of you had noticed, <a href=\"https://www.kaggle.com/competitions/image-matching-challenge-2024/discussion/487806#2733466\" target=\"_blank\">there were some inconsistencies</a> between some <code>image_name</code> entries in the <code>train_labels.csv</code> and the actual file names. I've posted a new version of the dataset that standardizes both the filenames and metadata to remove those inconsistencies.</p>\n<p>There are still some unlabeled image files in the train set; we're not planning on removing those.</p>\n<p>Apologies for the confusion and thank you all for your patience.</p>",
      "rawMarkdown": "As some of you had noticed, [there were some inconsistencies](https://www.kaggle.com/competitions/image-matching-challenge-2024/discussion/487806#2733466) between some `image_name` entries in the `train_labels.csv` and the actual file names. I've posted a new version of the dataset that standardizes both the filenames and metadata to remove those inconsistencies.\n\nThere are still some unlabeled image files in the train set; we're not planning on removing those.\n\nApologies for the confusion and thank you all for your patience.",
      "votes": null
    },
    {
      "id": "2735372",
      "postDate": "04/04/2024 17:30:25",
      "content": "<p>Please notice also that:</p>\n<ul>\n<li>In the training data <code>smf</code> sub-folders have been now correctly renamed as <code>sfm</code>;</li>\n<li>COLMAP 3D models are all included as .txt files.</li>\n</ul>\n<p>Thanks again Sohier!</p>",
      "rawMarkdown": "Please notice also that:\n- In the training data `smf` sub-folders have been now correctly renamed as `sfm`;\n- COLMAP 3D models are all included as .txt files.\n\nThanks again Sohier!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2735372,
      "author_name": "fabiobellavia",
      "author_url": "",
      "post_date": "04/04/2024 17:30:25",
      "content": "<p>Please notice also that:</p>\n<ul>\n<li>In the training data <code>smf</code> sub-folders have been now correctly renamed as <code>sfm</code>;</li>\n<li>COLMAP 3D models are all included as .txt files.</li>\n</ul>\n<p>Thanks again Sohier!</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2735342": "As some of you had noticed, [there were some inconsistencies](https://www.kaggle.com/competitions/image-matching-challenge-2024/discussion/487806#2733466) between some `image_name` entries in the `train_labels.csv` and the actual file names. I've posted a new version of the dataset that standardizes both the filenames and metadata to remove those inconsistencies.\n\nThere are still some unlabeled image files in the train set; we're not planning on removing those.\n\nApologies for the confusion and thank you all for your patience.",
    "2735372": "Please notice also that:\n- In the training data `smf` sub-folders have been now correctly renamed as `sfm`;\n- COLMAP 3D models are all included as .txt files.\n\nThanks again Sohier!"
  },
  "source": "meta"
}