{
  "id": 164189,
  "title": "CNN Architectures",
  "url": "/competitions/birdsong-recognition/discussion/164189",
  "author_name": "",
  "post_date": "2020-07-05T05:25:47.853219200Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>I am curious, for those who tried CNN, which architecture do you think works well in spectrogram classification, RESNET, DENSENET, INCEPTION or a simpler architecture? Also, how do you decide how to customize those layers since many combinations are possible?</p>",
  "messages": [
    {
      "id": "915805",
      "postDate": "07/05/2020 05:25:47",
      "content": "<p>I am curious, for those who tried CNN, which architecture do you think works well in spectrogram classification, RESNET, DENSENET, INCEPTION or a simpler architecture? Also, how do you decide how to customize those layers since many combinations are possible?</p>",
      "rawMarkdown": "I am curious, for those who tried CNN, which architecture do you think works well in spectrogram classification, RESNET, DENSENET, INCEPTION or a simpler architecture? Also, how do you decide how to customize those layers since many combinations are possible?",
      "votes": null
    },
    {
      "id": "916011",
      "postDate": "07/05/2020 09:23:45",
      "content": "<p>what about VGG16?</p>",
      "rawMarkdown": "what about VGG16?",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 916011,
      "author_name": "aboutakashsah",
      "author_url": "",
      "post_date": "07/05/2020 09:23:45",
      "content": "<p>what about VGG16?</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "915805": "I am curious, for those who tried CNN, which architecture do you think works well in spectrogram classification, RESNET, DENSENET, INCEPTION or a simpler architecture? Also, how do you decide how to customize those layers since many combinations are possible?",
    "916011": "what about VGG16?"
  },
  "source": "meta"
}