The Herbarium 2021 Half–Earth Challenge Dataset and Machine Learning Competition

Riccardo de Lutio; John Y. Park; Kimberly A. Watson; Stefano D'Aronco; Jan D. Wegner; Jan J. Wieringa; Melissa Tulig; Richard L. Pyle; Timothy J. Gallaher; Gillian Brown; Gordon Guymer; Andrew Franks; Dhahara Ranatunga; Yumiko Baba; Serge J. Belongie; Fabián A. Michelangeli; Barbara A. Ambrose; Damon P. Little

doi:10.3389/fpls.2021.787127

The Herbarium 2021 Half–Earth Challenge Dataset and Machine Learning Competition

Riccardo de Lutio^*, John Y. Park, Kimberly A. Watson, Stefano D'Aronco, Jan D. Wegner, Jan J. Wieringa, Melissa Tulig, Richard L. Pyle, Timothy J. Gallaher, Gillian Brown, Gordon Guymer, Andrew Franks, Dhahara Ranatunga, Yumiko Baba, Serge J. Belongie, Fabián A. Michelangeli, Barbara A. Ambrose, Damon P. Little

^*Corresponding author for this work

Image Analysis, Computational Modelling and Geometry

Research output: Contribution to journal › Journal article › Research › peer-review

7 Citations (Scopus)

33 Downloads (Pure)

Abstract

Herbarium sheets present a unique view of the world's botanical history, evolution, and biodiversity. This makes them an all–important data source for botanical research. With the increased digitization of herbaria worldwide and advances in the domain of fine–grained visual classification which can facilitate automatic identification of herbarium specimen images, there are many opportunities for supporting and expanding research in this field. However, existing datasets are either too small, or not diverse enough, in terms of represented taxa, geographic distribution, and imaging protocols. Furthermore, aggregating datasets is difficult as taxa are recognized under a multitude of names and must be aligned to a common reference. We introduce the Herbarium 2021 Half–Earth dataset: the largest and most diverse dataset of herbarium specimen images, to date, for automatic taxon recognition. We also present the results of the Herbarium 2021 Half–Earth challenge, a competition that was part of the Eighth Workshop on Fine-Grained Visual Categorization (FGVC8) and hosted by Kaggle to encourage the development of models to automatically identify taxa from herbarium sheet images.

Original language	English
Article number	787127
Journal	Frontiers in Plant Science
Volume	12
Pages (from-to)	1-15
ISSN	1664-462X
DOIs	https://doi.org/10.3389/fpls.2021.787127
Publication status	Published - 1 Feb 2022

Bibliographical note

Funding Information:
We would like to thank Kiat Chuan Tan from Google and the team at Kaggle (Walter Reade and Maggie Demkin) for their generous support in making this challenge possible. We would also like to thank everyone who entered the Herbarium 2021 Half?Earth Challenge. We are particularly grateful to the teams that provided detailed information on the model architectures and training strategies behind their winning submissions.

Funding Information:
This work was partially funded by National Science Foundation (USA) grant DEB 2054684.

Publisher Copyright:
Copyright © 2022 de Lutio, Park, Watson, D'Aronco, Wegner, Wieringa, Tulig, Pyle, Gallaher, Brown, Guymer, Franks, Ranatunga, Baba, Belongie, Michelangeli, Ambrose and Little.

Keywords

datasets
fine-grained visual categorization
herbarium specimen image
hierarchical classification
machine learning competition

Access to Document

10.3389/fpls.2021.787127Licence: CC BY

FulltextFinal published version, 3.81 MBLicence: CC BY

Cite this

de Lutio, R., Park, J. Y., Watson, K. A., D'Aronco, S., Wegner, J. D., Wieringa, J. J., Tulig, M., Pyle, R. L., Gallaher, T. J., Brown, G., Guymer, G., Franks, A., Ranatunga, D., Baba, Y., Belongie, S. J., Michelangeli, F. A., Ambrose, B. A., & Little, D. P. (2022). The Herbarium 2021 Half–Earth Challenge Dataset and Machine Learning Competition. Frontiers in Plant Science, 12, 1-15. Article 787127. https://doi.org/10.3389/fpls.2021.787127

de Lutio, R, Park, JY, Watson, KA, D'Aronco, S, Wegner, JD, Wieringa, JJ, Tulig, M, Pyle, RL, Gallaher, TJ, Brown, G, Guymer, G, Franks, A, Ranatunga, D, Baba, Y, Belongie, SJ, Michelangeli, FA, Ambrose, BA & Little, DP 2022, 'The Herbarium 2021 Half–Earth Challenge Dataset and Machine Learning Competition', Frontiers in Plant Science, vol. 12, 787127, pp. 1-15. https://doi.org/10.3389/fpls.2021.787127

@article{f0e393ede26b485ca6cc58af51bff01b,

title = "The Herbarium 2021 Half–Earth Challenge Dataset and Machine Learning Competition",

abstract = "Herbarium sheets present a unique view of the world's botanical history, evolution, and biodiversity. This makes them an all–important data source for botanical research. With the increased digitization of herbaria worldwide and advances in the domain of fine–grained visual classification which can facilitate automatic identification of herbarium specimen images, there are many opportunities for supporting and expanding research in this field. However, existing datasets are either too small, or not diverse enough, in terms of represented taxa, geographic distribution, and imaging protocols. Furthermore, aggregating datasets is difficult as taxa are recognized under a multitude of names and must be aligned to a common reference. We introduce the Herbarium 2021 Half–Earth dataset: the largest and most diverse dataset of herbarium specimen images, to date, for automatic taxon recognition. We also present the results of the Herbarium 2021 Half–Earth challenge, a competition that was part of the Eighth Workshop on Fine-Grained Visual Categorization (FGVC8) and hosted by Kaggle to encourage the development of models to automatically identify taxa from herbarium sheet images.",

keywords = "datasets, fine-grained visual categorization, herbarium specimen image, hierarchical classification, machine learning competition",

author = "{de Lutio}, Riccardo and Park, {John Y.} and Watson, {Kimberly A.} and Stefano D'Aronco and Wegner, {Jan D.} and Wieringa, {Jan J.} and Melissa Tulig and Pyle, {Richard L.} and Gallaher, {Timothy J.} and Gillian Brown and Gordon Guymer and Andrew Franks and Dhahara Ranatunga and Yumiko Baba and Belongie, {Serge J.} and Michelangeli, {Fabi{\'a}n A.} and Ambrose, {Barbara A.} and Little, {Damon P.}",

note = "Funding Information: We would like to thank Kiat Chuan Tan from Google and the team at Kaggle (Walter Reade and Maggie Demkin) for their generous support in making this challenge possible. We would also like to thank everyone who entered the Herbarium 2021 Half?Earth Challenge. We are particularly grateful to the teams that provided detailed information on the model architectures and training strategies behind their winning submissions. Funding Information: This work was partially funded by National Science Foundation (USA) grant DEB 2054684. Publisher Copyright: Copyright {\textcopyright} 2022 de Lutio, Park, Watson, D'Aronco, Wegner, Wieringa, Tulig, Pyle, Gallaher, Brown, Guymer, Franks, Ranatunga, Baba, Belongie, Michelangeli, Ambrose and Little.",

year = "2022",

month = feb,

day = "1",

doi = "10.3389/fpls.2021.787127",

language = "English",

volume = "12",

pages = "1--15",

journal = "Frontiers in Plant Science",

issn = "1664-462X",

publisher = "Frontiers Media S.A.",

}

TY - JOUR

T1 - The Herbarium 2021 Half–Earth Challenge Dataset and Machine Learning Competition

AU - de Lutio, Riccardo

AU - Park, John Y.

AU - Watson, Kimberly A.

AU - D'Aronco, Stefano

AU - Wegner, Jan D.

AU - Wieringa, Jan J.

AU - Tulig, Melissa

AU - Pyle, Richard L.

AU - Gallaher, Timothy J.

AU - Brown, Gillian

AU - Guymer, Gordon

AU - Franks, Andrew

AU - Ranatunga, Dhahara

AU - Baba, Yumiko

AU - Belongie, Serge J.

AU - Michelangeli, Fabián A.

AU - Ambrose, Barbara A.

AU - Little, Damon P.

N1 - Funding Information: We would like to thank Kiat Chuan Tan from Google and the team at Kaggle (Walter Reade and Maggie Demkin) for their generous support in making this challenge possible. We would also like to thank everyone who entered the Herbarium 2021 Half?Earth Challenge. We are particularly grateful to the teams that provided detailed information on the model architectures and training strategies behind their winning submissions. Funding Information: This work was partially funded by National Science Foundation (USA) grant DEB 2054684. Publisher Copyright: Copyright © 2022 de Lutio, Park, Watson, D'Aronco, Wegner, Wieringa, Tulig, Pyle, Gallaher, Brown, Guymer, Franks, Ranatunga, Baba, Belongie, Michelangeli, Ambrose and Little.

PY - 2022/2/1

Y1 - 2022/2/1

N2 - Herbarium sheets present a unique view of the world's botanical history, evolution, and biodiversity. This makes them an all–important data source for botanical research. With the increased digitization of herbaria worldwide and advances in the domain of fine–grained visual classification which can facilitate automatic identification of herbarium specimen images, there are many opportunities for supporting and expanding research in this field. However, existing datasets are either too small, or not diverse enough, in terms of represented taxa, geographic distribution, and imaging protocols. Furthermore, aggregating datasets is difficult as taxa are recognized under a multitude of names and must be aligned to a common reference. We introduce the Herbarium 2021 Half–Earth dataset: the largest and most diverse dataset of herbarium specimen images, to date, for automatic taxon recognition. We also present the results of the Herbarium 2021 Half–Earth challenge, a competition that was part of the Eighth Workshop on Fine-Grained Visual Categorization (FGVC8) and hosted by Kaggle to encourage the development of models to automatically identify taxa from herbarium sheet images.

AB - Herbarium sheets present a unique view of the world's botanical history, evolution, and biodiversity. This makes them an all–important data source for botanical research. With the increased digitization of herbaria worldwide and advances in the domain of fine–grained visual classification which can facilitate automatic identification of herbarium specimen images, there are many opportunities for supporting and expanding research in this field. However, existing datasets are either too small, or not diverse enough, in terms of represented taxa, geographic distribution, and imaging protocols. Furthermore, aggregating datasets is difficult as taxa are recognized under a multitude of names and must be aligned to a common reference. We introduce the Herbarium 2021 Half–Earth dataset: the largest and most diverse dataset of herbarium specimen images, to date, for automatic taxon recognition. We also present the results of the Herbarium 2021 Half–Earth challenge, a competition that was part of the Eighth Workshop on Fine-Grained Visual Categorization (FGVC8) and hosted by Kaggle to encourage the development of models to automatically identify taxa from herbarium sheet images.

KW - datasets

KW - fine-grained visual categorization

KW - herbarium specimen image

KW - hierarchical classification

KW - machine learning competition

UR - http://www.scopus.com/inward/record.url?scp=85124583035&partnerID=8YFLogxK

U2 - 10.3389/fpls.2021.787127

DO - 10.3389/fpls.2021.787127

M3 - Journal article

C2 - 35178056

AN - SCOPUS:85124583035

SN - 1664-462X

VL - 12

SP - 1

EP - 15

JO - Frontiers in Plant Science

JF - Frontiers in Plant Science

M1 - 787127

ER -

The Herbarium 2021 Half–Earth Challenge Dataset and Machine Learning Competition

Abstract

Bibliographical note

Keywords

Access to Document

Other files and links

Cite this