Oil Radish Dataset Semantic Segmentation and Yield Estimation Challenges

Organized by anderskm - Current server time: Jan. 18, 2021, 7:19 a.m. UTC


First phase
June 1, 2019, midnight UTC


Competition Ends

Oil Radish Growth Dataset Challenges

Welcome to the Oil Radish Growth Dataset challenges. This competitions consists of the challenges associated with the Oil Radish Growth Dataset presented at CVPPP 2019 workshop at CVPR 2019. The dataset was collected over a 7 week growing period of oil radish. It consists of pixel-wise annotated images with associated field measurements and unlabelled images of oil radish mixed with volunteer seeded barley, stubble, and weed. The dataset also contains weather data from the oil radish growing period.

The challenges associated with the dataset are the Semantic Segmentation challenge and the Yield Estimation challenge. In the Semantic Segmentation challenge, participants must perform pixel-wise classifiction on a subset of the labelled images. In the Yield Estimation challenge, participants must estimate the oil radish yield of same subset of labelled images.

For more details see the original dataset paper:

The Oil Radish Growth Dataset for Semantic Segmentation and Yield Estimation. A. K. Mortensen, S. Skovsen, H. Karstoft, and R. Gislum; The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2019

Semantic Segmentation challenge

The intersection over union (IoU) (or Jaccard index) is used as evaluation metric for the semantic segmentation challenge. For each submission, the per class, mean and class frequency weighted IoU are used as evaluation metrics.

Let c_{o,p} be the number of instances of class o predicted as class p, then the IoU for class k is given by:

IoU_k = \frac{c_{k,k}}{- c_{k,k} + \sum_{o=1}^K{c_{o,k}} + \sum_{p=1}^K{c_{k,p}}}

The mean IoU is then given by:

mean~IoU = \frac{1}{K}\sum_{k=1}^K{IoU_k}

And the class frequency weighted IoU is given by:

f.w.~IoU = \sum_{k=1}^K \left(IoU_k \frac{\sum_{o=1}^K{c_{o,k}}}{\sum_{o=1}^K\sum_{p=1}^K{c_{o,p}}} \right)

NOTE: Pixels annotated as "Unknown" are ignored during evaluation, but pixels predicted as "Unknown" are not.

Yield Estimation challenge

The yiels estimation challenge is evaluated using the root mean square error (RMSE) and the mean absolute percentage error (MAPE) for each field measurement: fresh weight, dry weight, C-content and N-content. The RMSE and MAPE are given by:

RMSE = \sqrt{\frac{1}{N}\sum_{i=1}^N{\left(t_i - y_i \right)^2}}


MAPE = \frac{1}{N}\sum_{i=1}^N{\left|\frac{t_i - y_i}{t_i}\right|100\%}

Copyright and license

Copyright: © 2019 Anders Krogh Mortensen, Aarhus University

The Oil Radish Growth Dataset is released for academic research only and is free to use for researchers from educational or research institutions for non-commercial purposes. When downloading the dataset, you agree to not reproduce, duplicate, sell, trade, resell or exploit for any commercial purposes, any parts of the dataset or parts derived from the dataset without explicit permission from the original authors. When using the dataset, please cite the original research paper associated with the dataset.

The data in the Oil Radish Growth Dataset and associated software (including but not limited to "starting kit" and "evalutation script") are provided "as is". When using the data and software associated with the dataset is at the users own risk. In no event shall the copyright holder or contributors be liable for any direct, indirect, incidental, special, exemplary, or consequential damages (including, but not limited to, procurement of substitute goods or services; loss of use, data, or profits; or buseness interruption) however caused and on any theory of liability, whether in contract, strict liability, or tort (including negligence or otherwise) arising in any way out of the use of this data, even if advised of the possibility of such damage.





Please cite the following paper, when using the Oil Radish Growth Dataset.

	author = {Krogh Mortensen, Anders and Skovsen, Soren and Karstoft, Henrik and Gislum, Rene},
	title = {The Oil Radish Growth Dataset for Semantic Segmentation and Yield Estimation},
	booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops},
	month = {June},
	year = {2019}

First phase

Start: June 1, 2019, midnight

Description: The competition only consists of one phase, which never ends. In this phase, partisipants can submit methods for both the semantic segmentation challenge and the yield estimation challenge. See the "Learn the Details" and "Participte" tabs for more information.

Competition Ends


You must be logged in to participate in competitions.

Sign In