SugarViT—Multi-objective regression of UAV images with Vision Transformers and Deep Label Distribution Learning demonstrated on disease severity prediction in sugar beet
PLOS ONE · 2025
DOI: 10.1371/journal.pone.0318097 ↗Kurz gefasst
Fachartikel in der Zeitschrift PLOS ONE, 2025. Verfasst von Maurice Günder und 5 weiteren Autor:innen. Beteiligte Einrichtung: Fraunhofer Institute for Intelligent Analysis and Information Systems IAIS, Schloss Birlinghoven, Sankt Augustin, Germany. Hauptthemen laut MEDLINE: Beta vulgaris, Plant Diseases, Remote Sensing Technology, Deep Learning. Die Arbeit wurde bisher 6-mal zitiert (Crossref) und nennt 43 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Deutsche Forschungsgemeinschaft, Federal Ministry of Education and Research of Germany and the state of North-Rhine Westphalia as part of the Lamarr-Institute for Machine Learning and Artifici….
Abstract (Originalsprache)
Remote sensing and artificial intelligence are pivotal technologies of precision agriculture nowadays. The efficient retrieval of large-scale field imagery combined with machine learning techniques shows success in various tasks like phenotyping, weeding, cropping, and disease control. This work will introduce a machine learning framework for automatized large-scale plant-specific trait annotation for the use case of disease severity scoring for CLS in sugar beet. With concepts of DLDL, special loss functions, and a tailored model architecture, we develop an efficient Vision Transformer based model for disease severity scoring called SugarViT. One novelty in this work is the combination of remote sensing data with environmental parameters of the experimental sites for disease severity prediction. Although the model is evaluated on this special use case, it is held as generic as possible to also be applicable to various image-based classification and regression tasks. With our framework, it is even possible to learn models on multi-objective problems, as we show by a pretraining on environmental metadata. Furthermore, we perform several comparison experiments with state-of-the-art methods and models to constitute our modeling and preprocessing choices.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Fraunhofer Institute for Intelligent Analysis and Information Systems IAIS, Schloss Birlinghoven, Sankt Augustin, Germany.
- Institute for Computer Science III, University of Bonn, Bonn, Germany.
- Institute for Sugar Beet Research (IfZ), Göttingen, Germany.
ORCID-Profile: Maurice Günder, Facundo Ramón Ispizua Yamati, Christian Bauckhage
Förderung
- Deutsche Forschungsgemeinschaft
- Federal Ministry of Education and Research of Germany and the state of North-Rhine Westphalia as part of the Lamarr-Institute for Machine Learning and Artifici…
Zitieren
APA
Günder, M., Yamati, F. R. I., Barreto, A., Mahlein, A. K., Sifa, R., & Bauckhage, C. (2025). SugarViT—Multi-objective regression of UAV images with Vision Transformers and Deep Label Distribution Learning demonstrated on disease severity prediction in sugar beet. PLOS ONE. https://doi.org/10.1371/journal.pone.0318097
Vancouver
Günder M, Yamati FRI, Barreto A, Mahlein AK, Sifa R, Bauckhage C. SugarViT—Multi-objective regression of UAV images with Vision Transformers and Deep Label Distribution Learning demonstrated on disease severity prediction in sugar beet. PLOS ONE. 2025. doi:10.1371/journal.pone.0318097
BibTeX
@article{Gnder2025,
author = {Günder, Maurice and Yamati, Facundo Ramón Ispizua and Barreto, Abel and Mahlein, Anne-Katrin and Sifa, Rafet and Bauckhage, Christian},
title = {SugarViT—Multi-objective regression of UAV images with Vision Transformers and Deep Label Distribution Learning demonstrated on disease severity prediction in sugar beet},
journal = {PLOS ONE},
year = {2025},
doi = {10.1371/journal.pone.0318097}
}
EUVIMED