{
  "id": 268157,
  "title": "Google Landmark Recognition 2019 Test Images + TFRecords",
  "url": "/competitions/landmark-recognition-2021/discussion/268157",
  "author_name": "",
  "post_date": "2021-08-26T08:22:20.366478400Z",
  "votes": 17,
  "comment_count": 1,
  "views": 0,
  "content": "<p>The Google Landmark Recognition 2019 competition did make the test dataset public, which can be downloaded using <a href=\"https://github.com/cvdfoundation/google-landmark\" target=\"_blank\">this</a> public GitHub code. This dataset gives a clear view of what we can expect from the test set. In 2019 98.3% of the test images were non-landmark, making non-landmark identification a key challenge in this competition. The dataset contains 117577 images, which are stored in their original size. The images in the TFReocords are downsized to size 384.</p>\n<p>The dataset can be found <a href=\"https://www.kaggle.com/markwijkhuizen/google-landmark-recognition-2019-test\" target=\"_blank\">here</a></p>\n<p>The notebook to create the dataset can be found <a href=\"https://www.kaggle.com/markwijkhuizen/google-landmark-recognition-2019-test-dataset-pub\" target=\"_blank\">here</a></p>",
  "messages": [
    {
      "id": "1491209",
      "postDate": "08/26/2021 08:22:20",
      "content": "<p>The Google Landmark Recognition 2019 competition did make the test dataset public, which can be downloaded using <a href=\"https://github.com/cvdfoundation/google-landmark\" target=\"_blank\">this</a> public GitHub code. This dataset gives a clear view of what we can expect from the test set. In 2019 98.3% of the test images were non-landmark, making non-landmark identification a key challenge in this competition. The dataset contains 117577 images, which are stored in their original size. The images in the TFReocords are downsized to size 384.</p>\n<p>The dataset can be found <a href=\"https://www.kaggle.com/markwijkhuizen/google-landmark-recognition-2019-test\" target=\"_blank\">here</a></p>\n<p>The notebook to create the dataset can be found <a href=\"https://www.kaggle.com/markwijkhuizen/google-landmark-recognition-2019-test-dataset-pub\" target=\"_blank\">here</a></p>",
      "rawMarkdown": "The Google Landmark Recognition 2019 competition did make the test dataset public, which can be downloaded using [this](https://github.com/cvdfoundation/google-landmark) public GitHub code. This dataset gives a clear view of what we can expect from the test set. In 2019 98.3% of the test images were non-landmark, making non-landmark identification a key challenge in this competition. The dataset contains 117577 images, which are stored in their original size. The images in the TFReocords are downsized to size 384.\n\nThe dataset can be found [here](https://www.kaggle.com/markwijkhuizen/google-landmark-recognition-2019-test)\n\nThe notebook to create the dataset can be found [here](https://www.kaggle.com/markwijkhuizen/google-landmark-recognition-2019-test-dataset-pub)",
      "votes": null
    },
    {
      "id": "1526228",
      "postDate": "09/28/2021 00:33:18",
      "content": "<p>Hi <a href=\"https://www.kaggle.com/markwijkhuizen\" target=\"_blank\">@markwijkhuizen</a>, thanks for creating this for us.</p>\n<p>Just wanna ask if you have tried attaching this 2019 test set for submission?<br>\nI did that and got an 'Notebook Threw Exception' error after submission.</p>\n<p>Is there anything i should take note of when using 2019 test set as part of the submission?<br>\nAppreciate any help from you.:)</p>",
      "rawMarkdown": "Hi @markwijkhuizen, thanks for creating this for us.\n\nJust wanna ask if you have tried attaching this 2019 test set for submission?\nI did that and got an 'Notebook Threw Exception' error after submission.\n\nIs there anything i should take note of when using 2019 test set as part of the submission?\nAppreciate any help from you.:)",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1526228,
      "author_name": "tmxxuan",
      "author_url": "",
      "post_date": "09/28/2021 00:33:18",
      "content": "<p>Hi <a href=\"https://www.kaggle.com/markwijkhuizen\" target=\"_blank\">@markwijkhuizen</a>, thanks for creating this for us.</p>\n<p>Just wanna ask if you have tried attaching this 2019 test set for submission?<br>\nI did that and got an 'Notebook Threw Exception' error after submission.</p>\n<p>Is there anything i should take note of when using 2019 test set as part of the submission?<br>\nAppreciate any help from you.:)</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1491209": "The Google Landmark Recognition 2019 competition did make the test dataset public, which can be downloaded using [this](https://github.com/cvdfoundation/google-landmark) public GitHub code. This dataset gives a clear view of what we can expect from the test set. In 2019 98.3% of the test images were non-landmark, making non-landmark identification a key challenge in this competition. The dataset contains 117577 images, which are stored in their original size. The images in the TFReocords are downsized to size 384.\n\nThe dataset can be found [here](https://www.kaggle.com/markwijkhuizen/google-landmark-recognition-2019-test)\n\nThe notebook to create the dataset can be found [here](https://www.kaggle.com/markwijkhuizen/google-landmark-recognition-2019-test-dataset-pub)",
    "1526228": "Hi @markwijkhuizen, thanks for creating this for us.\n\nJust wanna ask if you have tried attaching this 2019 test set for submission?\nI did that and got an 'Notebook Threw Exception' error after submission.\n\nIs there anything i should take note of when using 2019 test set as part of the submission?\nAppreciate any help from you.:)"
  },
  "source": "meta"
}