{
  "id": 308892,
  "title": "Anyone marked which photo Ids are not suitable for training?",
  "url": "/competitions/happy-whale-and-dolphin/discussion/308892",
  "author_name": "Anand Parthiban",
  "post_date": "2022-02-20T19:04:34.729000",
  "votes": 7,
  "comment_count": 2,
  "views": 0,
  "content": "<p>I've been seeing some EDA notebooks and posts showing a number of photos that have boats, people, multiple whales, etc in the photos. Has anyone created a CSV or noted down some annotations that mark these photo IDs in the dataset for data cleaning purposes? </p>",
  "messages": [
    {
      "id": 1698941,
      "postDate": "2022-02-20T19:04:34.730Z",
      "content": "<p>I've been seeing some EDA notebooks and posts showing a number of photos that have boats, people, multiple whales, etc in the photos. Has anyone created a CSV or noted down some annotations that mark these photo IDs in the dataset for data cleaning purposes? </p>",
      "rawMarkdown": "I've been seeing some EDA notebooks and posts showing a number of photos that have boats, people, multiple whales, etc in the photos. Has anyone created a CSV or noted down some annotations that mark these photo IDs in the dataset for data cleaning purposes? ",
      "votes": 7
    },
    {
      "id": 1698981,
      "postDate": "2022-02-20T19:58:32.343Z",
      "content": "<p>I spent 2 days watching almost 50% of dataset photos. All photos mentioned by you (with boats, people, multiple whales, etc.) are usable in this competition. There are other cases which nobody (in my opinion) is able to recognize whale ID (e.g. bieluga far far away … 15-30cm over the surface of water).  </p>",
      "rawMarkdown": "I spent 2 days watching almost 50% of dataset photos. All photos mentioned by you (with boats, people, multiple whales, etc.) are usable in this competition. There are other cases which nobody (in my opinion) is able to recognize whale ID (e.g. bieluga far far away … 15-30cm over the surface of water).  ",
      "votes": 1
    },
    {
      "id": 1699186,
      "postDate": "2022-02-21T01:55:53.583Z",
      "rawMarkdown": "",
      "isDeleted": true
    }
  ],
  "comments": [
    {
      "id": 1698981,
      "author_name": "Remek Kinas",
      "author_url": "",
      "post_date": "2022-02-20T19:58:32.343000",
      "content": "<p>I spent 2 days watching almost 50% of dataset photos. All photos mentioned by you (with boats, people, multiple whales, etc.) are usable in this competition. There are other cases which nobody (in my opinion) is able to recognize whale ID (e.g. bieluga far far away … 15-30cm over the surface of water).  </p>",
      "votes": 1,
      "replies": []
    },
    {
      "id": 1699186,
      "author_name": "",
      "author_url": "",
      "post_date": "2022-02-21T01:55:53.583000",
      "content": "",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1698941": "I've been seeing some EDA notebooks and posts showing a number of photos that have boats, people, multiple whales, etc in the photos. Has anyone created a CSV or noted down some annotations that mark these photo IDs in the dataset for data cleaning purposes? ",
    "1698981": "I spent 2 days watching almost 50% of dataset photos. All photos mentioned by you (with boats, people, multiple whales, etc.) are usable in this competition. There are other cases which nobody (in my opinion) is able to recognize whale ID (e.g. bieluga far far away … 15-30cm over the surface of water).  ",
    "1699186": ""
  }
}