{
  "id": 242551,
  "title": "EfficientNet: Improving Accuracy and Efficiency",
  "url": "/competitions/siim-covid19-detection/discussion/242551",
  "author_name": "",
  "post_date": "2021-05-29T16:06:42.182535700Z",
  "votes": 3,
  "comment_count": 3,
  "views": 0,
  "content": "<p>I have been working on this competition from past 1 week and I found that efficientnet is working good on this problem. So here are some important links i found :</p>\n<ol>\n<li><p><a href=\"url\" target=\"_blank\">https://ai.googleblog.com/2019/05/efficientnet-improving-accuracy-and.html</a></p></li>\n<li><p><a href=\"url\" target=\"_blank\">https://medium.com/analytics-vidhya/image-classification-with-efficientnet-better-performance-with-computational-efficiency-f480fdb00ac6</a></p></li>\n<li><p><a href=\"url\" target=\"_blank\">https://keras.io/api/applications/efficientnet/</a></p></li>\n</ol>",
  "messages": [
    {
      "id": "1327786",
      "postDate": "05/29/2021 16:06:42",
      "content": "<p>I have been working on this competition from past 1 week and I found that efficientnet is working good on this problem. So here are some important links i found :</p>\n<ol>\n<li><p><a href=\"url\" target=\"_blank\">https://ai.googleblog.com/2019/05/efficientnet-improving-accuracy-and.html</a></p></li>\n<li><p><a href=\"url\" target=\"_blank\">https://medium.com/analytics-vidhya/image-classification-with-efficientnet-better-performance-with-computational-efficiency-f480fdb00ac6</a></p></li>\n<li><p><a href=\"url\" target=\"_blank\">https://keras.io/api/applications/efficientnet/</a></p></li>\n</ol>",
      "rawMarkdown": "I have been working on this competition from past 1 week and I found that efficientnet is working good on this problem. So here are some important links i found :\n\n1. [https://ai.googleblog.com/2019/05/efficientnet-improving-accuracy-and.html](url)\n\n2. [https://medium.com/analytics-vidhya/image-classification-with-efficientnet-better-performance-with-computational-efficiency-f480fdb00ac6](url)\n\n3. [https://keras.io/api/applications/efficientnet/](url)",
      "votes": null
    },
    {
      "id": "1328160",
      "postDate": "05/30/2021 01:41:40",
      "content": "<p>yes even i have found efficient nets to be working well, which is surprising because architectures that are scalable and have been scaled on imagenet don't work well with radiology images. Inception V4 also gives very good results</p>",
      "rawMarkdown": "yes even i have found efficient nets to be working well, which is surprising because architectures that are scalable and have been scaled on imagenet don't work well with radiology images. Inception V4 also gives very good results",
      "votes": null
    },
    {
      "id": "1328198",
      "postDate": "05/30/2021 03:19:46",
      "content": "<p>yes, I was also surprised</p>",
      "rawMarkdown": "yes, I was also surprised",
      "votes": null
    },
    {
      "id": "1368442",
      "postDate": "06/28/2021 15:31:18",
      "content": "<p>Hello <a href=\"https://www.kaggle.com/varundutt9213\" target=\"_blank\">@varundutt9213</a> and <a href=\"https://www.kaggle.com/bhaveshkumar2806\" target=\"_blank\">@bhaveshkumar2806</a> I am new to this kind of competitions and I would like to learn more. Can you suggest some paper in which is explained how does networks like efficientnet perform on different type of datasets?</p>\n<p>Thanks in advance,</p>",
      "rawMarkdown": "Hello @varundutt9213 and @bhaveshkumar2806 I am new to this kind of competitions and I would like to learn more. Can you suggest some paper in which is explained how does networks like efficientnet perform on different type of datasets?\n\nThanks in advance,",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1328160,
      "author_name": "varundutt9213",
      "author_url": "",
      "post_date": "05/30/2021 01:41:40",
      "content": "<p>yes even i have found efficient nets to be working well, which is surprising because architectures that are scalable and have been scaled on imagenet don't work well with radiology images. Inception V4 also gives very good results</p>",
      "votes": null,
      "replies": [
        {
          "id": 1328198,
          "author_name": "bhaveshkumar2806",
          "author_url": "",
          "post_date": "05/30/2021 03:19:46",
          "content": "<p>yes, I was also surprised</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1368442,
      "author_name": "saraserafini1",
      "author_url": "",
      "post_date": "06/28/2021 15:31:18",
      "content": "<p>Hello <a href=\"https://www.kaggle.com/varundutt9213\" target=\"_blank\">@varundutt9213</a> and <a href=\"https://www.kaggle.com/bhaveshkumar2806\" target=\"_blank\">@bhaveshkumar2806</a> I am new to this kind of competitions and I would like to learn more. Can you suggest some paper in which is explained how does networks like efficientnet perform on different type of datasets?</p>\n<p>Thanks in advance,</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1327786": "I have been working on this competition from past 1 week and I found that efficientnet is working good on this problem. So here are some important links i found :\n\n1. [https://ai.googleblog.com/2019/05/efficientnet-improving-accuracy-and.html](url)\n\n2. [https://medium.com/analytics-vidhya/image-classification-with-efficientnet-better-performance-with-computational-efficiency-f480fdb00ac6](url)\n\n3. [https://keras.io/api/applications/efficientnet/](url)",
    "1328160": "yes even i have found efficient nets to be working well, which is surprising because architectures that are scalable and have been scaled on imagenet don't work well with radiology images. Inception V4 also gives very good results",
    "1328198": "yes, I was also surprised",
    "1368442": "Hello @varundutt9213 and @bhaveshkumar2806 I am new to this kind of competitions and I would like to learn more. Can you suggest some paper in which is explained how does networks like efficientnet perform on different type of datasets?\n\nThanks in advance,"
  },
  "source": "meta"
}