{
  "id": 159290,
  "title": "Pre-processing and Augmentation libraries?",
  "url": "/competitions/birdsong-recognition/discussion/159290",
  "author_name": "Krishnendu Sengupta",
  "post_date": "2020-06-17T02:24:24.239000",
  "votes": 1,
  "comment_count": 3,
  "views": 0,
  "content": "<p>What tools do you use for audio processing for Deep Learning tasks? Is there something like albumentations for audio?</p>",
  "messages": [
    {
      "id": 889596,
      "postDate": "2020-06-17T03:13:39.360Z",
      "content": "<ul>\n<li>torchaudio (<a href=\"https://pytorch.org/audio/transforms.html\">https://pytorch.org/audio/transforms.html</a>)</li>\n<li>pysox (<a href=\"https://github.com/rabitt/pysox\">https://github.com/rabitt/pysox</a>)</li>\n<li>nlpaug (<a href=\"https://github.com/makcedward/nlpaug\">https://github.com/makcedward/nlpaug</a>)</li>\n<li>pysndfx (<a href=\"https://github.com/carlthome/python-audio-effects\">https://github.com/carlthome/python-audio-effects</a>)</li>\n<li>pyroomacoustics (<a href=\"https://github.com/LCAV/pyroomacoustics\">https://github.com/LCAV/pyroomacoustics</a>)</li>\n</ul>\n\n<p>No golden standard as far as I know.</p>\n\n<p>Note that pysox and pysndfx relies on sox (<a href=\"http://sox.sourceforge.net/\">http://sox.sourceforge.net/</a>), therefore the basic function is the same. You can also use sox directly to get augmented sound clips.</p>",
      "rawMarkdown": "* torchaudio (https://pytorch.org/audio/transforms.html)\n* pysox (https://github.com/rabitt/pysox)\n* nlpaug (https://github.com/makcedward/nlpaug)\n* pysndfx (https://github.com/carlthome/python-audio-effects)\n* pyroomacoustics (https://github.com/LCAV/pyroomacoustics)\n\nNo golden standard as far as I know.\n\nNote that pysox and pysndfx relies on sox (http://sox.sourceforge.net/), therefore the basic function is the same. You can also use sox directly to get augmented sound clips.",
      "votes": 4,
      "replies": [
        {
          "id": 889727,
          "postDate": "2020-06-17T05:21:20.750Z",
          "rawMarkdown": "",
          "isDeleted": true
        }
      ]
    },
    {
      "id": 889557,
      "postDate": "2020-06-17T02:24:24.240Z",
      "content": "<p>What tools do you use for audio processing for Deep Learning tasks? Is there something like albumentations for audio?</p>",
      "rawMarkdown": "What tools do you use for audio processing for Deep Learning tasks? Is there something like albumentations for audio?",
      "votes": 1
    },
    {
      "id": 988215,
      "postDate": "2020-08-27T22:09:35.130Z",
      "content": "<p>Hey this will help <a href=\"https://www.kaggle.com/tanulsingh077/audio-albumentations-transform-your-audio\" target=\"_blank\">https://www.kaggle.com/tanulsingh077/audio-albumentations-transform-your-audio</a></p>",
      "rawMarkdown": "Hey this will help https://www.kaggle.com/tanulsingh077/audio-albumentations-transform-your-audio"
    }
  ],
  "comments": [
    {
      "id": 889596,
      "author_name": "Hidehisa Arai",
      "author_url": "",
      "post_date": "2020-06-17T03:13:39.360000",
      "content": "<ul>\n<li>torchaudio (<a href=\"https://pytorch.org/audio/transforms.html\">https://pytorch.org/audio/transforms.html</a>)</li>\n<li>pysox (<a href=\"https://github.com/rabitt/pysox\">https://github.com/rabitt/pysox</a>)</li>\n<li>nlpaug (<a href=\"https://github.com/makcedward/nlpaug\">https://github.com/makcedward/nlpaug</a>)</li>\n<li>pysndfx (<a href=\"https://github.com/carlthome/python-audio-effects\">https://github.com/carlthome/python-audio-effects</a>)</li>\n<li>pyroomacoustics (<a href=\"https://github.com/LCAV/pyroomacoustics\">https://github.com/LCAV/pyroomacoustics</a>)</li>\n</ul>\n\n<p>No golden standard as far as I know.</p>\n\n<p>Note that pysox and pysndfx relies on sox (<a href=\"http://sox.sourceforge.net/\">http://sox.sourceforge.net/</a>), therefore the basic function is the same. You can also use sox directly to get augmented sound clips.</p>",
      "votes": 4,
      "replies": [
        {
          "id": 889727,
          "author_name": "",
          "author_url": "",
          "post_date": "2020-06-17T05:21:20.750000",
          "content": "",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 988215,
      "author_name": "Mr_KnowNothing",
      "author_url": "",
      "post_date": "2020-08-27T22:09:35.130000",
      "content": "<p>Hey this will help <a href=\"https://www.kaggle.com/tanulsingh077/audio-albumentations-transform-your-audio\" target=\"_blank\">https://www.kaggle.com/tanulsingh077/audio-albumentations-transform-your-audio</a></p>",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "889596": "* torchaudio (https://pytorch.org/audio/transforms.html)\n* pysox (https://github.com/rabitt/pysox)\n* nlpaug (https://github.com/makcedward/nlpaug)\n* pysndfx (https://github.com/carlthome/python-audio-effects)\n* pyroomacoustics (https://github.com/LCAV/pyroomacoustics)\n\nNo golden standard as far as I know.\n\nNote that pysox and pysndfx relies on sox (http://sox.sourceforge.net/), therefore the basic function is the same. You can also use sox directly to get augmented sound clips.",
    "889557": "What tools do you use for audio processing for Deep Learning tasks? Is there something like albumentations for audio?",
    "988215": "Hey this will help https://www.kaggle.com/tanulsingh077/audio-albumentations-transform-your-audio"
  }
}