{
  "id": 266256,
  "title": "Can landmarks in the train set appear in the test set?",
  "url": "/competitions/landmark-retrieval-2021/discussion/266256",
  "author_name": "",
  "post_date": "2021-08-18T13:03:23.759100700Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Question in title. I was reading the descriptions of the data, and it didn't mention anything about landmarks in the train set reappearing in the test set. Did anyone find such cases?</p>",
  "messages": [
    {
      "id": "1479407",
      "postDate": "08/18/2021 13:03:23",
      "content": "<p>Question in title. I was reading the descriptions of the data, and it didn't mention anything about landmarks in the train set reappearing in the test set. Did anyone find such cases?</p>",
      "rawMarkdown": "Question in title. I was reading the descriptions of the data, and it didn't mention anything about landmarks in the train set reappearing in the test set. Did anyone find such cases?",
      "votes": null
    },
    {
      "id": "1481941",
      "postDate": "08/19/2021 18:39:25",
      "content": "<p>Hi,<br>\nSome details are mentioned in the 'Data Description' part of Recognition competition:<br>\n<a href=\"https://www.kaggle.com/c/landmark-recognition-2021/data\" target=\"_blank\">https://www.kaggle.com/c/landmark-recognition-2021/data</a></p>\n<p>\"To facilitate recognition-by-retrieval approaches, the private training set contains only a 100k subset of the total public training set. This 100k subset contains all of the training set images associated with the landmarks in the private test set. You may still attach the full training set as an external data set if you wish.\"</p>\n<p>Particularly, considering the close overlap of test data from both competitions as someone observed:<br>\n<a href=\"https://www.kaggle.com/inoueu1/landmark-retrieval-recognition-data-compare\" target=\"_blank\">https://www.kaggle.com/inoueu1/landmark-retrieval-recognition-data-compare</a></p>",
      "rawMarkdown": "Hi,\nSome details are mentioned in the 'Data Description' part of Recognition competition:\nhttps://www.kaggle.com/c/landmark-recognition-2021/data\n\n\"To facilitate recognition-by-retrieval approaches, the private training set contains only a 100k subset of the total public training set. This 100k subset contains all of the training set images associated with the landmarks in the private test set. You may still attach the full training set as an external data set if you wish.\"\n\nParticularly, considering the close overlap of test data from both competitions as someone observed:\nhttps://www.kaggle.com/inoueu1/landmark-retrieval-recognition-data-compare",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1481941,
      "author_name": "krishvanapalli",
      "author_url": "",
      "post_date": "08/19/2021 18:39:25",
      "content": "<p>Hi,<br>\nSome details are mentioned in the 'Data Description' part of Recognition competition:<br>\n<a href=\"https://www.kaggle.com/c/landmark-recognition-2021/data\" target=\"_blank\">https://www.kaggle.com/c/landmark-recognition-2021/data</a></p>\n<p>\"To facilitate recognition-by-retrieval approaches, the private training set contains only a 100k subset of the total public training set. This 100k subset contains all of the training set images associated with the landmarks in the private test set. You may still attach the full training set as an external data set if you wish.\"</p>\n<p>Particularly, considering the close overlap of test data from both competitions as someone observed:<br>\n<a href=\"https://www.kaggle.com/inoueu1/landmark-retrieval-recognition-data-compare\" target=\"_blank\">https://www.kaggle.com/inoueu1/landmark-retrieval-recognition-data-compare</a></p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1479407": "Question in title. I was reading the descriptions of the data, and it didn't mention anything about landmarks in the train set reappearing in the test set. Did anyone find such cases?",
    "1481941": "Hi,\nSome details are mentioned in the 'Data Description' part of Recognition competition:\nhttps://www.kaggle.com/c/landmark-recognition-2021/data\n\n\"To facilitate recognition-by-retrieval approaches, the private training set contains only a 100k subset of the total public training set. This 100k subset contains all of the training set images associated with the landmarks in the private test set. You may still attach the full training set as an external data set if you wish.\"\n\nParticularly, considering the close overlap of test data from both competitions as someone observed:\nhttps://www.kaggle.com/inoueu1/landmark-retrieval-recognition-data-compare"
  },
  "source": "meta"
}