# 2018 FGCVx Fungi Classification Challenge — Data

来源：https://www.kaggle.com/competitions/fungi-challenge-fgvc-2018/data

本地 Meta Kaggle 快照；不是本次实时页面。

# Dataset Description

<h2>Data</h2>
Download the images and annotations [here](https://github.com/visipedia/fgvcx_fungi_comp#data).
<h2>File Descriptions</h2>
<ul>
<li><strong>fungi_train_val.tgz</strong> - The training and validation images in a directory structure following `images/category/image.jpg`</li>
<li><strong>fungi_test.tgz</strong> - The testing images in a single directory `test/`</li>
<li><strong>train_val_annotations.tgz</strong> - Contains training and validation annotations.</li>
<li><strong>test_information.tgz</strong> - Contains testing image information.</li>
<li><strong>fungi_sample_submission.csv</strong> - A sample submission file in the correct format.</li>
</ul>
<h2>Image Format</h2>
<p>All images have been saved in the JPEG format and have been resized to have a maximum dimension of 1024 pixels.</p>

<h2>Annotation Format</h2>

We closely follow the annotation format of the [COCO dataset][2]. The annotations are stored in the [JSON][3] format and are organized as follows:
<pre>
{
  "info" : info,
  "images" : [image],
  "categories" : [category],
  "annotations" : [annotation],
  "licenses" : [license]
}

info{
  "year" : int,
  "version" : str,
  "description" : str,
  "contributor" : str,
  "url" : str,
  "date_created" : datetime,
}

image{
  "id" : int,
  "width" : int,
  "height" : int,
  "file_name" : str,
  "license" : int,
  "rights_holder" : str
}

category{
  "id" : int,
  "name" : str,
  "supercategory" : str,
}

annotation{
  "id" : int,
  "image_id" : int,
  "category_id" : int
}

license{
  "id" : int,
  "name" : str,
  "url" : str
}`
</pre>

  [2]: http://cocodataset.org/#download
  [3]: http://www.json.org/