{
  "id": 47320,
  "title": "[Video] Gold medal solution by Sergey Mushinsky (Eng subtitles)",
  "url": "/competitions/carvana-image-masking-challenge/discussion/47320",
  "author_name": "",
  "post_date": "2018-01-12T02:06:28.081293Z",
  "votes": 23,
  "comment_count": 1,
  "views": 0,
  "content": "<p>In case you missed it:</p>\n\n<p><a href=\"https://www.youtube.com/watch?v=bNJLfEkqU0c\">https://www.youtube.com/watch?v=bNJLfEkqU0c</a></p>\n\n<p>Sergey Mushinskiy tells his team's solution of Kaggle Carvana Image Masking Challenge. In this competition, Carvana is challenging Kagglers to develop an algorithm that automatically removes the photo studio background. From this video you learn:</p>\n\n<ul>\n<li><p>Using pseudo-labeling to train network with predicted test samples</p></li>\n<li><p>Teamwork with shared GPUs time Is it useful to hand label</p></li>\n<li>Other approaches: from two network without any averaging to complex ensemble with insane architectures</li>\n</ul>",
  "messages": [
    {
      "id": "267666",
      "postDate": "01/12/2018 02:06:28",
      "content": "<p>In case you missed it:</p>\n\n<p><a href=\"https://www.youtube.com/watch?v=bNJLfEkqU0c\">https://www.youtube.com/watch?v=bNJLfEkqU0c</a></p>\n\n<p>Sergey Mushinskiy tells his team's solution of Kaggle Carvana Image Masking Challenge. In this competition, Carvana is challenging Kagglers to develop an algorithm that automatically removes the photo studio background. From this video you learn:</p>\n\n<ul>\n<li><p>Using pseudo-labeling to train network with predicted test samples</p></li>\n<li><p>Teamwork with shared GPUs time Is it useful to hand label</p></li>\n<li>Other approaches: from two network without any averaging to complex ensemble with insane architectures</li>\n</ul>",
      "rawMarkdown": "In case you missed it:\n\nhttps://www.youtube.com/watch?v=bNJLfEkqU0c\n\nSergey Mushinskiy tells his team's solution of Kaggle Carvana Image Masking Challenge. In this competition, Carvana is challenging Kagglers to develop an algorithm that automatically removes the photo studio background. From this video you learn:\n\n - Using pseudo-labeling to train network with predicted test samples\n   \n - Teamwork with shared GPUs time Is it useful to hand label\n - Other approaches: from two network without any averaging to complex ensemble with insane architectures",
      "votes": null
    },
    {
      "id": "1912310",
      "postDate": "08/24/2022 16:33:49",
      "content": "<p>Thank u so much!</p>",
      "rawMarkdown": "Thank u so much!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1912310,
      "author_name": "dangluong",
      "author_url": "",
      "post_date": "08/24/2022 16:33:49",
      "content": "<p>Thank u so much!</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "267666": "In case you missed it:\n\nhttps://www.youtube.com/watch?v=bNJLfEkqU0c\n\nSergey Mushinskiy tells his team's solution of Kaggle Carvana Image Masking Challenge. In this competition, Carvana is challenging Kagglers to develop an algorithm that automatically removes the photo studio background. From this video you learn:\n\n - Using pseudo-labeling to train network with predicted test samples\n   \n - Teamwork with shared GPUs time Is it useful to hand label\n - Other approaches: from two network without any averaging to complex ensemble with insane architectures",
    "1912310": "Thank u so much!"
  },
  "source": "meta"
}