{
  "id": 151783,
  "title": "augmentation techniques",
  "url": "/competitions/global-wheat-detection/discussion/151783",
  "author_name": "",
  "post_date": "2020-05-17T03:18:57.771141200Z",
  "votes": 26,
  "comment_count": 5,
  "views": 0,
  "content": "<p>I have published a notebook with amazing augmentation techniques <a href=\"https://www.kaggle.com/nvnnghia/awesome-augmentation?scriptVersionId=34144701\" target=\"_blank\">here</a></p>\n<p>By playing with this augmentation resources and some tweakings. I can get upto 0.72 LB with a single Faster RCNN (no pseudo labeling).  </p>",
  "messages": [
    {
      "id": "850804",
      "postDate": "05/17/2020 03:18:57",
      "content": "<p>I have published a notebook with amazing augmentation techniques <a href=\"https://www.kaggle.com/nvnnghia/awesome-augmentation?scriptVersionId=34144701\" target=\"_blank\">here</a></p>\n<p>By playing with this augmentation resources and some tweakings. I can get upto 0.72 LB with a single Faster RCNN (no pseudo labeling).  </p>",
      "rawMarkdown": "I have published a notebook with amazing augmentation techniques [here](https://www.kaggle.com/nvnnghia/awesome-augmentation?scriptVersionId=34144701)\n\nBy playing with this augmentation resources and some tweakings. I can get upto 0.72 LB with a single Faster RCNN (no pseudo labeling).",
      "votes": null
    },
    {
      "id": "850863",
      "postDate": "05/17/2020 05:07:25",
      "content": "<p>Good job! Thank you for your sharing. I've tested much data augmentation techs, but few have improved LB.</p>",
      "rawMarkdown": "Good job! Thank you for your sharing. I've tested much data augmentation techs, but few have improved LB.",
      "votes": null
    },
    {
      "id": "851130",
      "postDate": "05/17/2020 11:22:42",
      "content": "<p>Thank you very much</p>",
      "rawMarkdown": "Thank you very much",
      "votes": null
    },
    {
      "id": "892535",
      "postDate": "06/19/2020 01:46:24",
      "content": "<p>sorry，I am a newbie in competition, what is LB？</p>",
      "rawMarkdown": "sorry，I am a newbie in competition, what is LB？",
      "votes": null
    },
    {
      "id": "892553",
      "postDate": "06/19/2020 02:12:53",
      "content": "<p><a href=\"/lalalayujian\">@lalalayujian</a> it means public LeaderBoad score</p>",
      "rawMarkdown": "lalalayujian it means public LeaderBoad score",
      "votes": null
    },
    {
      "id": "895022",
      "postDate": "06/21/2020 03:43:48",
      "content": "<p><a href=\"/nvnnghia\">@nvnnghia</a> hi there, thanks for your sharing. Did you used script in the hyperlink to get LB 0.72? I used functions in that kernel but can't reach to 0.67 with a single FasterRCNN model. </p>",
      "rawMarkdown": "nvnnghia hi there, thanks for your sharing. Did you used script in the hyperlink to get LB 0.72? I used functions in that kernel but can't reach to 0.67 with a single FasterRCNN model.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 850863,
      "author_name": "whurobin",
      "author_url": "",
      "post_date": "05/17/2020 05:07:25",
      "content": "<p>Good job! Thank you for your sharing. I've tested much data augmentation techs, but few have improved LB.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 851130,
      "author_name": "imvivek14",
      "author_url": "",
      "post_date": "05/17/2020 11:22:42",
      "content": "<p>Thank you very much</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 892535,
      "author_name": "lalalayujian",
      "author_url": "",
      "post_date": "06/19/2020 01:46:24",
      "content": "<p>sorry，I am a newbie in competition, what is LB？</p>",
      "votes": null,
      "replies": [
        {
          "id": 892553,
          "author_name": "nvnnghia",
          "author_url": "",
          "post_date": "06/19/2020 02:12:53",
          "content": "<p><a href=\"/lalalayujian\">@lalalayujian</a> it means public LeaderBoad score</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 895022,
      "author_name": "",
      "author_url": "",
      "post_date": "06/21/2020 03:43:48",
      "content": "<p><a href=\"/nvnnghia\">@nvnnghia</a> hi there, thanks for your sharing. Did you used script in the hyperlink to get LB 0.72? I used functions in that kernel but can't reach to 0.67 with a single FasterRCNN model. </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "850804": "I have published a notebook with amazing augmentation techniques [here](https://www.kaggle.com/nvnnghia/awesome-augmentation?scriptVersionId=34144701)\n\nBy playing with this augmentation resources and some tweakings. I can get upto 0.72 LB with a single Faster RCNN (no pseudo labeling).",
    "850863": "Good job! Thank you for your sharing. I've tested much data augmentation techs, but few have improved LB.",
    "851130": "Thank you very much",
    "892535": "sorry，I am a newbie in competition, what is LB？",
    "892553": "lalalayujian it means public LeaderBoad score",
    "895022": "nvnnghia hi there, thanks for your sharing. Did you used script in the hyperlink to get LB 0.72? I used functions in that kernel but can't reach to 0.67 with a single FasterRCNN model."
  },
  "source": "meta"
}