April 12, 2021 Dataset Open Access

Quentin Geissmann

Citation Style Language JSON Export

{
  "publisher": "Zenodo", 
  "DOI": "10.5281/zenodo.4680119", 
  "language": "eng", 
  "title": "Sticky Pi -- Machine Learning Data, Configuration and Models", 
  "issued": {
    "date-parts": [
      [
        2021, 
        4, 
        12
      ]
    ]
  }, 
  "abstract": "<p><strong>Dataset for the Machine Learning section of the Sticky Pi project (https://doc.sticky-pi.com/)</strong></p>\n\n<p>Contains the dataset for the three algorithms described in the publication: Universal Insect Detector, Siamese Insect Matcher and Insect Tuboid Classifier.</p>\n\n<p><strong>Universal Insect Detector:</strong></p>\n\n<p>`universal_insect_detector/` contains training/validation data, configuration files to train the model, and the model as trained and used for publication.</p>\n\n<ul>\n\t<li>`data/` &ndash; A set of svg images that contain the embedded jpg raw image, and a set of non-intersecting polygon around the labelled insects</li>\n\t<li>`output/`\n\t<ul>\n\t\t<li>`model_final.pth` &ndash; the model as trained for the publication</li>\n\t</ul>\n\t</li>\n\t<li>`config/`\n\t<ul>\n\t\t<li>`config.yaml` &ndash; The configuration file defining the hyperparameters to train the model as well as the taxonomic labels</li>\n\t\t<li>`config.yaml `&ndash; The configuration file defining the hyperparameters to train the model</li>\n\t\t<li>`mask_rcnn_R_101_C4_3x.yaml` &ndash; the base configuration file from which config is derived</li>\n\t</ul>\n\t</li>\n</ul>\n\n<p>&nbsp;</p>\n\n<p><strong>Siamese Insect Matcher</strong></p>\n\n<p>`siamese_insect_matcher/` contains training/validation data, configuration files to train the model, and the model as trained and used for publication.</p>\n\n<ul>\n\t<li>`data/` &ndash; a set of svg images that contain two embedded jpg raw images vertically stacked corresponding to two frames in a series. Each predicted insect is labelled as a polygon. Insects that are labelled as the same instance, between the two frames, are grouped (i.e. SVG group). The filename of each image is `&lt;device&gt;.&lt;datetime_frame_1&gt;.&lt;datetime_frame_2&gt;.svg`</li>\n\t<li>`output/`\n\t<ul>\n\t\t<li>`model_final.pth` &ndash; the model as trained for the publication</li>\n\t</ul>\n\t</li>\n\t<li>`config/`\n\t<ul>\n\t\t<li>`config.yaml` &ndash; The configuration file defining the hyperparameters to train the model as well as the taxonomic labels</li>\n\t\t<li>`config.yaml` &ndash; The configuration file defining the hyperparameters to train the model</li>\n\t</ul>\n\t</li>\n</ul>\n\n<p>&nbsp;</p>\n\n<p><strong>Insect Tuboid Classifier:</strong></p>\n\n<p>`insect_tuboid_classifier/` contains images of insect tuboid, a database file describing their taxonomy, a configuration file to train the model, and the model as trained and used for publication.</p>\n\n<ul>\n\t<li>`data/`\n\t<ul>\n\t\t<li>`database.db`: a sqlite file with a single table `ANNOTATIONS`. The table maps a unique identifier of each tuboid (tuboid_id) to a set of manually annotated taxonomic variables.</li>\n\t\t<li>A directory tree of the form: `&lt;series_id&gt;/&lt;tuboid_id&gt;/`. Each terminal directory contains:\n\t\t<ul>\n\t\t\t<li>\n\t\t\t<ul>\n\t\t\t\t<li>`tuboid.jpg` &ndash; a jpeg image made of 224 x 224 tiles representing all the shots in a tuboid, left to right, top to bottom &ndash; might be padded with empty images</li>\n\t\t\t\t<li>`metadata.txt` &ndash; a csv text file with columns:\n\t\t\t\t<ul>\n\t\t\t\t\t<li>\n\t\t\t\t\t<ul>\n\t\t\t\t\t\t<li>parrent_image_id &ndash; &lt;device&gt;.&lt;UTC_datetime&gt;</li>\n\t\t\t\t\t\t<li>X &ndash; the X coordinates of the object centroid</li>\n\t\t\t\t\t\t<li>Y &ndash; the Y coordinates of the object centroid</li>\n\t\t\t\t\t</ul>\n\t\t\t\t\t</li>\n\t\t\t\t</ul>\n\t\t\t\t</li>\n\t\t\t\t<li>scale &ndash; The scaling factor applied between the original and image and the 224 x 224 tile (&gt;1 =&gt; image was enlarged)</li>\n\t\t\t\t<li>`context.jpg` &ndash; a representation of the first whole image of a series, with a box around the first tuboid shot (this is for debugging/labelling purposes)</li>\n\t\t\t</ul>\n\t\t\t</li>\n\t\t</ul>\n\t\t</li>\n\t</ul>\n\t</li>\n\t<li>`output/`\n\t<ul>\n\t\t<li>`model_final.pth` &ndash; the model as trained for the publication</li>\n\t</ul>\n\t</li>\n\t<li>config/\n\t<ul>\n\t\t<li>`config.yaml` &ndash; The configuration file defining the hyperparameters to train the model as well as the taxonomic labels</li>\n\t</ul>\n\t</li>\n</ul>", 
  "author": [
    {
      "family": "Quentin Geissmann"
    }
  ], 
  "type": "dataset", 
  "id": "4680119"
}