Find file History
Permalink
Type Name Latest commit message Commit time
..
Failed to load latest commit information.
assets moving 2017 to its own dir Jan 27, 2018
svgs small text tweaks regarding the boxes Apr 10, 2018
README.md small text tweaks regarding the boxes Apr 10, 2018
README_raw.md small text tweaks regarding the boxes Apr 10, 2018

README.md

Banner

iNaturalist Competition

Please open an issue if you have questions or problems with the dataset.

2017 Competition

The 2017 competition, sponsored by Google, is part of the FGVC^4 workshop at CVPR.

Kaggle

We are using Kaggle to host the leaderboard. Checkout the competition page here.

Dates

Data Released April 5, 2017
Submission Server Open June 1, 2017
Submission Deadline July 7, 2017
Winners Announced July 21, 2017

Details

There are a total of 5,089 categories in the dataset, with 579,184 training images and 95,986 validation images. For the training set, the distribution of images per category follows the observation frequency of that category by the iNaturalist community. Therefore, there is a non-uniform distribution of images per category. Example images, along with their unique GBIF ID numbers (where available), can be viewed here.

Super Category Category Count Train Images Val Images
Plantae 2,101 158,407 38,206
Insecta 1,021 100,479 18,076
Aves 964 214,295 21,226
Reptilia 289 35,201 5,680
Mammalia 186 29,333 3,490
Fungi 121 5,826 1,780
Amphibia 115 15,318 2,385
Mollusca 93 7,536 1,841
Animalia 77 5,228 1,362
Arachnida 56 4,873 1,086
Actinopterygii 53 1,982 637
Chromista 9 398 144
Protozoa 4 308 73
Total 5,089 579,184 95,986

Evaluation

We follow a similar metric to the classification tasks of the ILSVRC. For each image , an algorithm will produce 5 labels , . We allow 5 labels because some categories are disambiguated with additional data provided by the observer, such as latitude, longitude and date. It might also be the case that multiple categories occur in an image (e.g. a photo of a bee on a flower). For this competition each image has one ground truth label , and the error for that image is:

Where

The overall error score for an algorithm is the average error over all test images:

Guidelines

Participants are restricted to train their algorithms on iNaturalist 2017 train and validation sets. Pretrained models may be used to construct the algorithms (e.g. ImageNet pretrained models) as long as participants do not actively collect additional data for the target categories of the iNaturalist 2017 competition. Please specify any and all external data used for training when uploading results.

The general rule is that we want participants to use only the provided training and validation images to train a model to classify the test images. We do not want participants crawling the web in search of additional data for the target categories. Participants should be in the mindset that this is the only data available for those categories.

Participants are allowed to collect additional annotations (e.g. bounding boxes) on the provided training and validation sets. Teams should specify that they collected additional annotations when submitting results.

Annotation Format

We closely follow the annotation format of the COCO dataset. The annotations are stored in the JSON format and are organized as follows:

{
  "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
}

Submission Format

The submission format for the Kaggle competition is a csv file with the following format:

id,predicted
12345,0 78 23 3 42
67890,83 13 42 0 21

The id column corresponds to the test image id. The predicted column corresponds to 5 category ids, separated by spaces. You should have one row for each test image.

Bounding Boxes

A subset of the dataset has been annotated with bounding boxes, please see the paper for the full details on how the boxes were collected. The following super categories were annotated:

Super Category Train Boxes Val Boxes
Insecta 106,304 16,732
Aves 283,931 17,314
Reptilia 36,476 5,480
Mammalia 31,109 2,654
Amphibia 15,812 2,280
Mollusca 8,566 1,571
Animalia 6,643 1,143
Arachnida 4,752 1,051
Actinopterygii 2,571 511
Total 496,164 48,736

The bounding box format follows the COCO format:

annotation{
  "id" : int,
  "image_id" : int,
  "category_id" : int,
  "bbox" : [x, y, width, height],
  "area" : float,
  "iscrowd" : 0
}

The bbox units are in pixels, the origin is the upper left hand corner, and the area value is approximated as (width * height) / 2.0 since we did not collect segmentation masks.

Bounding Box Caveats:

  • Crowdworkers were asked to annotate at the super category level rather than the category level. Therefore, for images with multiple box annotations there is no guarantee that all instances actually belong to the same category even though they are labeled as being the same category (e.g. multiple bird species could be boxed in an image and labeled as the same species). Rather, the boxed instances will belong to the same super category. Due to this problem, the validation set is restricted to images with single instances so that we can be confident the category label is correct for the given box.
  • iscrowd is hard coded to 0 for all annotations, although it is possible that a box is around a crowd of instances (such as barnacles or mussels).

Terms of Use

By downloading this dataset you agree to the following terms:

  1. You will abide by the iNaturalist Terms of Service
  2. You will use the data only for non-commercial research and educational purposes.
  3. You will NOT distribute the above images.
  4. The California Institute of Technology makes no representations or warranties regarding the data, including but not limited to warranties of non-infringement or fitness for a particular purpose.
  5. You accept full responsibility for your use of the data and shall defend and indemnify the California Institute of Technology, including its employees, officers and agents, against any and all claims arising from your use of the data, including but not limited to your use of any copies of copyrighted images that you may create from the data.

Data

Download the dataset files here: