A deep-learning framework for enhancing habitat identification
based on species composition

J 2024

A deep-learning framework for enhancing habitat identification based on species composition

LEBLANC, Cesar; Pierre BONNET; Maximilien SERVAJEAN; Milan CHYTRÝ; Svetlana ACIC et. al.

Basic information

Original name

A deep-learning framework for enhancing habitat identification based on species composition

Authors

LEBLANC, Cesar (guarantor); Pierre BONNET; Maximilien SERVAJEAN; Milan CHYTRÝ (203 Czech Republic, belonging to the institution); Svetlana ACIC; Olivier ARGAGNON; Ariel BERGAMINI; Idoia BIURRUN; Gianmaria BONARI; Juan A CAMPOS; Andraz CARNI; Renata CUSTEREVSKA; De Sanctis MICHELE; Juergen DENGLER; Emmanuel GARBOLINO; Valentin GOLUB; Ute JANDT; Florian JANSEN; Maria LEBEDEVA; Jonathan LENOIR; Jesper Erenskjold MOESLUND; Aaron PEREZ-HAASE; Remigiusz PIELECH; Jozef SIBIK; Zvjezdana STANCIC; Angela STANISCI; Grzegorz SWACHA; Domas UOGINTAS; Kiril VASSILEV; Thomas WOHLGEMUTH and Alexis JOLY

Edition

Applied Vegetation Science, HOBOKEN, Wiley-Blackwell, 2024, 1402-2001

Other information

Language

English

Type of outcome

Article in a journal

Field of Study

10611 Plant sciences, botany

Country of publisher

United States of America

Confidentiality degree

is not subject to a state or trade secret

References:

URL

Impact factor

Impact factor: 2.000 in 2023

RIV identification code

RIV/00216224:14310/24:00138040

Organization unit

Faculty of Science

DOI

http://dx.doi.org/10.1111/avsc.12802

UT WoS

001298764100001

EID Scopus

2-s2.0-85202490569

Keywords in English

artificial intelligence; biodiversity monitoring; deep learning; European flora; expert system; habitat type identification; phytosociology; species composition; vascular plants; vegetation classification

Abstract

In the original language

AimsThe accurate classification of habitats is essential for effective biodiversity conservation. The goal of this study was to harness the potential of deep learning to advance habitat identification in Europe. We aimed to develop and evaluate models capable of assigning vegetation-plot records to the habitats of the European Nature Information System (EUNIS), a widely used reference framework for European habitat types.LocationThe framework was designed for use in Europe and adjacent areas (e.g., Anatolia, Caucasus).MethodsWe leveraged deep-learning techniques, such as transformers (i.e., models with attention components able to learn contextual relations between categorical and numerical features) that we trained using spatial k-fold cross-validation (CV) on vegetation plots sourced from the European Vegetation Archive (EVA), to show that they have great potential for classifying vegetation-plot records. We tested different network architectures, feature encodings, hyperparameter tuning and noise addition strategies to identify the optimal model. We used an independent test set from the National Plant Monitoring Scheme (NPMS) to evaluate its performance and compare its results against the traditional expert systems.ResultsExploration of the use of deep learning applied to species composition and plot-location criteria for habitat classification led to the development of a framework containing a wide range of models. Our selected algorithm, applied to European habitat types, significantly improved habitat classification accuracy, achieving a more than twofold improvement compared to the previous state-of-the-art (SOTA) method on an external data set, clearly outperforming expert systems. The framework is shared and maintained through a GitHub repository.ConclusionsOur results demonstrate the potential benefits of the adoption of deep learning for improving the accuracy of vegetation classification. They highlight the importance of incorporating advanced technologies into habitat monitoring. These algorithms have shown to be better suited for habitat type prediction than expert systems. They push the accuracy score on a database containing hundreds of thousands of standardized presence/absence European surveys to 88.74%, as assessed by expert judgment. Finally, our results showcase that species dominance is a strong marker of ecosystems and that the exact cover abundance of the flora is not required to train neural networks with predictive performances. The framework we developed can be used by researchers and practitioners to accurately classify habitats.

Cite

LEBLANC, Cesar; Pierre BONNET; Maximilien SERVAJEAN; Milan CHYTRÝ; Svetlana ACIC; Olivier ARGAGNON; Ariel BERGAMINI; Idoia BIURRUN; Gianmaria BONARI; Juan A CAMPOS; Andraz CARNI; Renata CUSTEREVSKA; De Sanctis MICHELE; Juergen DENGLER; Emmanuel GARBOLINO; Valentin GOLUB; Ute JANDT; Florian JANSEN; Maria LEBEDEVA; Jonathan LENOIR; Jesper Erenskjold MOESLUND; Aaron PEREZ-HAASE; Remigiusz PIELECH; Jozef SIBIK; Zvjezdana STANCIC; Angela STANISCI; Grzegorz SWACHA; Domas UOGINTAS; Kiril VASSILEV; Thomas WOHLGEMUTH and Alexis JOLY. A deep-learning framework for enhancing habitat identification based on species composition. Applied Vegetation Science. HOBOKEN: Wiley-Blackwell, 2024, vol. 27, No 3, p. "e12802", 17 pp. ISSN 1402-2001. Available from: https://dx.doi.org/10.1111/avsc.12802.

@article{2460825,
   author = {Leblanc, Cesar and Bonnet, Pierre and Servajean, Maximilien and Chytrý, Milan and Acic, Svetlana and Argagnon, Olivier and Bergamini, Ariel and Biurrun, Idoia and Bonari, Gianmaria and Campos, Juan A and Carni, Andraz and Custerevska, Renata and Michele, De Sanctis and Dengler, Juergen and Garbolino, Emmanuel and Golub, Valentin and Jandt, Ute and Jansen, Florian and Lebedeva, Maria and Lenoir, Jonathan and Moeslund, Jesper Erenskjold and PerezandHaase, Aaron and Pielech, Remigiusz and Sibik, Jozef and Stancic, Zvjezdana and Stanisci, Angela and Swacha, Grzegorz and Uogintas, Domas and Vassilev, Kiril and Wohlgemuth, Thomas and Joly, Alexis},
   article_location = {HOBOKEN},
   article_number = {3},
   doi = {http://dx.doi.org/10.1111/avsc.12802},
   keywords = {artificial intelligence; biodiversity monitoring; deep learning; European flora; expert system; habitat type identification; phytosociology; species composition; vascular plants; vegetation classification},
   language = {eng},
   issn = {1402-2001},
   journal = {Applied Vegetation Science},
   title = {A deep-learning framework for enhancing habitat identification based on species composition},
   url = {https://doi.org/10.1111/avsc.12802},
   volume = {27},
   year = {2024}
}

TY  - JOUR
ID  - 2460825
AU  - Leblanc, Cesar - Bonnet, Pierre - Servajean, Maximilien - Chytrý, Milan - Acic, Svetlana - Argagnon, Olivier - Bergamini, Ariel - Biurrun, Idoia - Bonari, Gianmaria - Campos, Juan A - Carni, Andraz - Custerevska, Renata - Michele, De Sanctis - Dengler, Juergen - Garbolino, Emmanuel - Golub, Valentin - Jandt, Ute - Jansen, Florian - Lebedeva, Maria - Lenoir, Jonathan - Moeslund, Jesper Erenskjold - Perez-Haase, Aaron - Pielech, Remigiusz - Sibik, Jozef - Stancic, Zvjezdana - Stanisci, Angela - Swacha, Grzegorz - Uogintas, Domas - Vassilev, Kiril - Wohlgemuth, Thomas - Joly, Alexis
PY  - 2024
TI  - A deep-learning framework for enhancing habitat identification based on species composition
JF  - Applied Vegetation Science
VL  - 27
IS  - 3
SP  - "e12802"
EP  - "e12802"
PB  - Wiley-Blackwell
SN  - 14022001
KW  - artificial intelligence
KW  - biodiversity monitoring
KW  - deep learning
KW  - European flora
KW  - expert system
KW  - habitat type identification
KW  - phytosociology
KW  - species composition
KW  - vascular plants
KW  - vegetation classification
UR  - https://doi.org/10.1111/avsc.12802
N2  - AimsThe accurate classification of habitats is essential for effective biodiversity conservation. The goal of this study was to harness the potential of deep learning to advance habitat identification in Europe. We aimed to develop and evaluate models capable of assigning vegetation-plot records to the habitats of the European Nature Information System (EUNIS), a widely used reference framework for European habitat types.LocationThe framework was designed for use in Europe and adjacent areas (e.g., Anatolia, Caucasus).MethodsWe leveraged deep-learning techniques, such as transformers (i.e., models with attention components able to learn contextual relations between categorical and numerical features) that we trained using spatial k-fold cross-validation (CV) on vegetation plots sourced from the European Vegetation Archive (EVA), to show that they have great potential for classifying vegetation-plot records. We tested different network architectures, feature encodings, hyperparameter tuning and noise addition strategies to identify the optimal model. We used an independent test set from the National Plant Monitoring Scheme (NPMS) to evaluate its performance and compare its results against the traditional expert systems.ResultsExploration of the use of deep learning applied to species composition and plot-location criteria for habitat classification led to the development of a framework containing a wide range of models. Our selected algorithm, applied to European habitat types, significantly improved habitat classification accuracy, achieving a more than twofold improvement compared to the previous state-of-the-art (SOTA) method on an external data set, clearly outperforming expert systems. The framework is shared and maintained through a GitHub repository.ConclusionsOur results demonstrate the potential benefits of the adoption of deep learning for improving the accuracy of vegetation classification. They highlight the importance of incorporating advanced technologies into habitat monitoring. These algorithms have shown to be better suited for habitat type prediction than expert systems. They push the accuracy score on a database containing hundreds of thousands of standardized presence/absence European surveys to 88.74%, as assessed by expert judgment. Finally, our results showcase that species dominance is a strong marker of ecosystems and that the exact cover abundance of the flora is not required to train neural networks with predictive performances. The framework we developed can be used by researchers and practitioners to accurately classify habitats.
ER  -

LEBLANC, Cesar; Pierre BONNET; Maximilien SERVAJEAN; Milan CHYTRÝ; Svetlana ACIC; Olivier ARGAGNON; Ariel BERGAMINI; Idoia BIURRUN; Gianmaria BONARI; Juan A CAMPOS; Andraz CARNI; Renata CUSTEREVSKA; De Sanctis MICHELE; Juergen DENGLER; Emmanuel GARBOLINO; Valentin GOLUB; Ute JANDT; Florian JANSEN; Maria LEBEDEVA; Jonathan LENOIR; Jesper Erenskjold MOESLUND; Aaron PEREZ-HAASE; Remigiusz PIELECH; Jozef SIBIK; Zvjezdana STANCIC; Angela STANISCI; Grzegorz SWACHA; Domas UOGINTAS; Kiril VASSILEV; Thomas WOHLGEMUTH and Alexis JOLY. A deep-learning framework for enhancing habitat identification based on species composition. \textit{Applied Vegetation Science}. HOBOKEN: Wiley-Blackwell, 2024, vol.~27, No~3, p.~''e12802'', 17 pp. ISSN~1402-2001. Available from: https://dx.doi.org/10.1111/avsc.12802.

Přehled o publikaci