{
  "id": 254745,
  "title": "All you need to catchup quickly to SIIM-FISABIO-RSNA",
  "url": "/competitions/siim-covid19-detection/discussion/254745",
  "author_name": "Dr. Amritpal Singh",
  "post_date": "2021-07-23T13:24:09.169000",
  "votes": 24,
  "comment_count": 0,
  "views": 0,
  "content": "<p>I recently joined this competition, and am curating a list of resources to catch up. It's a bit late to join, but looking forward to studying the best solutions provided by the amazing Kaggle community.</p>\n<h2>Data augmentation</h2>\n<ul>\n<li>Simple Techniques:<ul>\n<li>Random SunFlare,    Random Fog,     Random Brightness/Contrast,  Random Crop,     Random Gamma,  HorizontalFlip/VerticalFlip,  Random Contrast,  Blur, and its varients<br>\nAffine,    Channel Dropout,    Inverting Image</li></ul></li>\n<li>Complex Techniques:<ul>\n<li>Random Erase,     Cut Out Augmentation ,     Cut Mix Augmentation ,     Mixup Augmentation ,     Mosaic Augmentation ,     Copy Paste Augmentation ,     Hide and Seek,     Grid Masking</li></ul></li>\n</ul>\n<h3>Most frequent Architectures being used -</h3>\n<ul>\n<li>EfficientNet b7</li>\n<li>CascadeRCNN - <a href=\"https://www.kaggle.com/sreevishnudamodaran/siim-effnetv2-l-cascadercnn-mmdetection-infer\" target=\"_blank\">link</a> by <a href=\"https://www.kaggle.com/sreevishnudamodaran\" target=\"_blank\">@sreevishnudamodaran</a></li>\n<li>YoloV5  - <a href=\"https://www.kaggle.com/ayuraj/train-covid-19-detection-using-yolov5\" target=\"_blank\">link</a></li>\n<li>Detectron2 - <a href=\"https://www.kaggle.com/ammarnassanalhajali/siim-covid-19-detectron2-training\" target=\"_blank\">link</a> by <a href=\"https://www.kaggle.com/ammarnassanalhajali\" target=\"_blank\">@ammarnassanalhajali</a></li>\n</ul>\n<h2>Highest scored notebooks -</h2>\n<ul>\n<li>SIIM EffNetV2_L CascadeRCNN MMDetection Infer  by <a href=\"https://www.kaggle.com/0sreevishnudamodaran\" target=\"_blank\">@0sreevishnudamodaran</a> <a href=\"https://www.kaggle.com/sreevishnudamodaran/siim-effnetv2-l-cascadercnn-mmdetection-infer\" target=\"_blank\">link</a> - score - 0.611</li>\n</ul>\n<h2>Interesting discussions</h2>\n<ul>\n<li>Metrics explained - <a href=\"https://www.kaggle.com/c/siim-covid19-detection/discussion/253345\" target=\"_blank\">link</a></li>\n<li>External dataset -<a href=\"https://www.kaggle.com/c/siim-covid19-detection/discussion/254160\" target=\"_blank\">link</a></li>\n<li>Links to resized dataset - <a href=\"https://www.kaggle.com/c/siim-covid19-detection/discussion/239918\" target=\"_blank\">link</a></li>\n<li>Winning solutions from past competitions - <a href=\"https://www.kaggle.com/c/siim-covid19-detection/discussion/239898\" target=\"_blank\">link</a></li>\n<li>Golden starter Kit - <a href=\"https://www.kaggle.com/c/siim-covid19-detection/discussion/240233\" target=\"_blank\">link</a> </li>\n</ul>\n<h3>Input Image size used</h3>\n<ul>\n<li>256</li>\n<li>512</li>\n</ul>\n<h3>Batch size</h3>\n<ul>\n<li>32</li>\n<li>16</li>\n</ul>\n<h4>(Work In Progress)</h4>\n<h4>I would love suggestions on other good notebooks and discussions to try that I have missed.</h4>\n<p>I hope this is helpful to the Kaggle community.</p>",
  "messages": [
    {
      "id": 1397757,
      "postDate": "2021-07-23T13:24:09.170Z",
      "content": "<p>I recently joined this competition, and am curating a list of resources to catch up. It's a bit late to join, but looking forward to studying the best solutions provided by the amazing Kaggle community.</p>\n<h2>Data augmentation</h2>\n<ul>\n<li>Simple Techniques:<ul>\n<li>Random SunFlare,    Random Fog,     Random Brightness/Contrast,  Random Crop,     Random Gamma,  HorizontalFlip/VerticalFlip,  Random Contrast,  Blur, and its varients<br>\nAffine,    Channel Dropout,    Inverting Image</li></ul></li>\n<li>Complex Techniques:<ul>\n<li>Random Erase,     Cut Out Augmentation ,     Cut Mix Augmentation ,     Mixup Augmentation ,     Mosaic Augmentation ,     Copy Paste Augmentation ,     Hide and Seek,     Grid Masking</li></ul></li>\n</ul>\n<h3>Most frequent Architectures being used -</h3>\n<ul>\n<li>EfficientNet b7</li>\n<li>CascadeRCNN - <a href=\"https://www.kaggle.com/sreevishnudamodaran/siim-effnetv2-l-cascadercnn-mmdetection-infer\" target=\"_blank\">link</a> by <a href=\"https://www.kaggle.com/sreevishnudamodaran\" target=\"_blank\">@sreevishnudamodaran</a></li>\n<li>YoloV5  - <a href=\"https://www.kaggle.com/ayuraj/train-covid-19-detection-using-yolov5\" target=\"_blank\">link</a></li>\n<li>Detectron2 - <a href=\"https://www.kaggle.com/ammarnassanalhajali/siim-covid-19-detectron2-training\" target=\"_blank\">link</a> by <a href=\"https://www.kaggle.com/ammarnassanalhajali\" target=\"_blank\">@ammarnassanalhajali</a></li>\n</ul>\n<h2>Highest scored notebooks -</h2>\n<ul>\n<li>SIIM EffNetV2_L CascadeRCNN MMDetection Infer  by <a href=\"https://www.kaggle.com/0sreevishnudamodaran\" target=\"_blank\">@0sreevishnudamodaran</a> <a href=\"https://www.kaggle.com/sreevishnudamodaran/siim-effnetv2-l-cascadercnn-mmdetection-infer\" target=\"_blank\">link</a> - score - 0.611</li>\n</ul>\n<h2>Interesting discussions</h2>\n<ul>\n<li>Metrics explained - <a href=\"https://www.kaggle.com/c/siim-covid19-detection/discussion/253345\" target=\"_blank\">link</a></li>\n<li>External dataset -<a href=\"https://www.kaggle.com/c/siim-covid19-detection/discussion/254160\" target=\"_blank\">link</a></li>\n<li>Links to resized dataset - <a href=\"https://www.kaggle.com/c/siim-covid19-detection/discussion/239918\" target=\"_blank\">link</a></li>\n<li>Winning solutions from past competitions - <a href=\"https://www.kaggle.com/c/siim-covid19-detection/discussion/239898\" target=\"_blank\">link</a></li>\n<li>Golden starter Kit - <a href=\"https://www.kaggle.com/c/siim-covid19-detection/discussion/240233\" target=\"_blank\">link</a> </li>\n</ul>\n<h3>Input Image size used</h3>\n<ul>\n<li>256</li>\n<li>512</li>\n</ul>\n<h3>Batch size</h3>\n<ul>\n<li>32</li>\n<li>16</li>\n</ul>\n<h4>(Work In Progress)</h4>\n<h4>I would love suggestions on other good notebooks and discussions to try that I have missed.</h4>\n<p>I hope this is helpful to the Kaggle community.</p>",
      "rawMarkdown": "I recently joined this competition, and am curating a list of resources to catch up. It's a bit late to join, but looking forward to studying the best solutions provided by the amazing Kaggle community.\n\n## Data augmentation \n- Simple Techniques:\n  - Random SunFlare,    Random Fog,     Random Brightness/Contrast,  Random Crop,     Random Gamma,  HorizontalFlip/VerticalFlip,  Random Contrast,  Blur, and its varients\n    Affine,    Channel Dropout,    Inverting Image\n- Complex Techniques:\n  - Random Erase,     Cut Out Augmentation ,     Cut Mix Augmentation ,     Mixup Augmentation ,     Mosaic Augmentation ,     Copy Paste Augmentation ,     Hide and Seek,     Grid Masking\n\n\n### Most frequent Architectures being used -\n- EfficientNet b7\n- CascadeRCNN - [link](https://www.kaggle.com/sreevishnudamodaran/siim-effnetv2-l-cascadercnn-mmdetection-infer) by @sreevishnudamodaran\n- YoloV5  - [link](https://www.kaggle.com/ayuraj/train-covid-19-detection-using-yolov5)\n- Detectron2 - [link](https://www.kaggle.com/ammarnassanalhajali/siim-covid-19-detectron2-training) by @ammarnassanalhajali\n\n\n## Highest scored notebooks - \n- SIIM EffNetV2_L CascadeRCNN MMDetection Infer  by @0sreevishnudamodaran [link](https://www.kaggle.com/sreevishnudamodaran/siim-effnetv2-l-cascadercnn-mmdetection-infer) - score - 0.611\n\n## Interesting discussions\n- Metrics explained - [link](https://www.kaggle.com/c/siim-covid19-detection/discussion/253345)\n- External dataset -[link](https://www.kaggle.com/c/siim-covid19-detection/discussion/254160)\n- Links to resized dataset - [link](https://www.kaggle.com/c/siim-covid19-detection/discussion/239918)\n- Winning solutions from past competitions - [link](https://www.kaggle.com/c/siim-covid19-detection/discussion/239898)\n- Golden starter Kit - [link](https://www.kaggle.com/c/siim-covid19-detection/discussion/240233) \n\n### Input Image size used\n- 256\n- 512\n\n###  Batch size \n- 32\n- 16\n\n#### (Work In Progress)\n#### I would love suggestions on other good notebooks and discussions to try that I have missed. \nI hope this is helpful to the Kaggle community.\n",
      "votes": 20
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "1397757": "I recently joined this competition, and am curating a list of resources to catch up. It's a bit late to join, but looking forward to studying the best solutions provided by the amazing Kaggle community.\n\n## Data augmentation \n- Simple Techniques:\n  - Random SunFlare,    Random Fog,     Random Brightness/Contrast,  Random Crop,     Random Gamma,  HorizontalFlip/VerticalFlip,  Random Contrast,  Blur, and its varients\n    Affine,    Channel Dropout,    Inverting Image\n- Complex Techniques:\n  - Random Erase,     Cut Out Augmentation ,     Cut Mix Augmentation ,     Mixup Augmentation ,     Mosaic Augmentation ,     Copy Paste Augmentation ,     Hide and Seek,     Grid Masking\n\n\n### Most frequent Architectures being used -\n- EfficientNet b7\n- CascadeRCNN - [link](https://www.kaggle.com/sreevishnudamodaran/siim-effnetv2-l-cascadercnn-mmdetection-infer) by @sreevishnudamodaran\n- YoloV5  - [link](https://www.kaggle.com/ayuraj/train-covid-19-detection-using-yolov5)\n- Detectron2 - [link](https://www.kaggle.com/ammarnassanalhajali/siim-covid-19-detectron2-training) by @ammarnassanalhajali\n\n\n## Highest scored notebooks - \n- SIIM EffNetV2_L CascadeRCNN MMDetection Infer  by @0sreevishnudamodaran [link](https://www.kaggle.com/sreevishnudamodaran/siim-effnetv2-l-cascadercnn-mmdetection-infer) - score - 0.611\n\n## Interesting discussions\n- Metrics explained - [link](https://www.kaggle.com/c/siim-covid19-detection/discussion/253345)\n- External dataset -[link](https://www.kaggle.com/c/siim-covid19-detection/discussion/254160)\n- Links to resized dataset - [link](https://www.kaggle.com/c/siim-covid19-detection/discussion/239918)\n- Winning solutions from past competitions - [link](https://www.kaggle.com/c/siim-covid19-detection/discussion/239898)\n- Golden starter Kit - [link](https://www.kaggle.com/c/siim-covid19-detection/discussion/240233) \n\n### Input Image size used\n- 256\n- 512\n\n###  Batch size \n- 32\n- 16\n\n#### (Work In Progress)\n#### I would love suggestions on other good notebooks and discussions to try that I have missed. \nI hope this is helpful to the Kaggle community.\n"
  }
}