BERT Bi-modal self-supervised learning for crop classification using Sentinel-2 and Planetscope
Crop identification and monitoring of crop dynamics are essential for agricultural planning, environmental monitoring, and ensuring food security. Recent advancements in remote sensing technology and state-of-the-art machine learning have enabled large-scale automated crop classification. However, t...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2025-05-01
|
| Series: | Frontiers in Remote Sensing |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/frsen.2025.1555887/full |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850282539284955136 |
|---|---|
| author | Ankit Patnala Martin G. Schultz Martin G. Schultz Juergen Gall Juergen Gall |
| author_facet | Ankit Patnala Martin G. Schultz Martin G. Schultz Juergen Gall Juergen Gall |
| author_sort | Ankit Patnala |
| collection | DOAJ |
| description | Crop identification and monitoring of crop dynamics are essential for agricultural planning, environmental monitoring, and ensuring food security. Recent advancements in remote sensing technology and state-of-the-art machine learning have enabled large-scale automated crop classification. However, these methods rely on labeled training data, which requires skilled human annotators or extensive field campaigns, making the process expensive and time-consuming. Self-supervised learning techniques have demonstrated promising results in leveraging large unlabeled datasets across domains. Yet, self-supervised representation learning for crop classification from remote sensing time series remains under-explored due to challenges in curating suitable pretext tasks. While bimodal self-supervised approaches combining data from Sentinel-2 and Planetscope sensors have facilitated pre-training, existing methods primarily exploit the distinct spectral properties of these complementary data sources. In this work, we propose novel self-supervised pre-training strategies inspired from BERT that leverage both the spectral and temporal resolution of Sentinel-2 and Planetscope imagery. We carry out extensive experiments comparing our approach to existing baseline setups across nine test cases, in which our method outperforms the baselines in eight instances. This pre-training thus offers an effective representation of crops for tasks such as crop classification. |
| format | Article |
| id | doaj-art-b67275d694344d2bb9f46b57d1ea74fb |
| institution | OA Journals |
| issn | 2673-6187 |
| language | English |
| publishDate | 2025-05-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Remote Sensing |
| spelling | doaj-art-b67275d694344d2bb9f46b57d1ea74fb2025-08-20T01:47:58ZengFrontiers Media S.A.Frontiers in Remote Sensing2673-61872025-05-01610.3389/frsen.2025.15558871555887BERT Bi-modal self-supervised learning for crop classification using Sentinel-2 and PlanetscopeAnkit Patnala0Martin G. Schultz1Martin G. Schultz2Juergen Gall3Juergen Gall4Juelich Supercomputing Centre, Forschungszentrum Juelich, Juelich, GermanyJuelich Supercomputing Centre, Forschungszentrum Juelich, Juelich, GermanyDepartment of Mathematics and Computer Science, University of Cologne, Cologne, GermanyDepartment of Information Systems and Artificial Intelligence, University of Bonn, Bonn, GermanyLamarr Institute for Machine Learning and Artificial Intelligence, Dortmund, GermanyCrop identification and monitoring of crop dynamics are essential for agricultural planning, environmental monitoring, and ensuring food security. Recent advancements in remote sensing technology and state-of-the-art machine learning have enabled large-scale automated crop classification. However, these methods rely on labeled training data, which requires skilled human annotators or extensive field campaigns, making the process expensive and time-consuming. Self-supervised learning techniques have demonstrated promising results in leveraging large unlabeled datasets across domains. Yet, self-supervised representation learning for crop classification from remote sensing time series remains under-explored due to challenges in curating suitable pretext tasks. While bimodal self-supervised approaches combining data from Sentinel-2 and Planetscope sensors have facilitated pre-training, existing methods primarily exploit the distinct spectral properties of these complementary data sources. In this work, we propose novel self-supervised pre-training strategies inspired from BERT that leverage both the spectral and temporal resolution of Sentinel-2 and Planetscope imagery. We carry out extensive experiments comparing our approach to existing baseline setups across nine test cases, in which our method outperforms the baselines in eight instances. This pre-training thus offers an effective representation of crops for tasks such as crop classification.https://www.frontiersin.org/articles/10.3389/frsen.2025.1555887/fullBERTbi-modal contrastive learningself-supervised learningremote sensingcrop classification |
| spellingShingle | Ankit Patnala Martin G. Schultz Martin G. Schultz Juergen Gall Juergen Gall BERT Bi-modal self-supervised learning for crop classification using Sentinel-2 and Planetscope Frontiers in Remote Sensing BERT bi-modal contrastive learning self-supervised learning remote sensing crop classification |
| title | BERT Bi-modal self-supervised learning for crop classification using Sentinel-2 and Planetscope |
| title_full | BERT Bi-modal self-supervised learning for crop classification using Sentinel-2 and Planetscope |
| title_fullStr | BERT Bi-modal self-supervised learning for crop classification using Sentinel-2 and Planetscope |
| title_full_unstemmed | BERT Bi-modal self-supervised learning for crop classification using Sentinel-2 and Planetscope |
| title_short | BERT Bi-modal self-supervised learning for crop classification using Sentinel-2 and Planetscope |
| title_sort | bert bi modal self supervised learning for crop classification using sentinel 2 and planetscope |
| topic | BERT bi-modal contrastive learning self-supervised learning remote sensing crop classification |
| url | https://www.frontiersin.org/articles/10.3389/frsen.2025.1555887/full |
| work_keys_str_mv | AT ankitpatnala bertbimodalselfsupervisedlearningforcropclassificationusingsentinel2andplanetscope AT martingschultz bertbimodalselfsupervisedlearningforcropclassificationusingsentinel2andplanetscope AT martingschultz bertbimodalselfsupervisedlearningforcropclassificationusingsentinel2andplanetscope AT juergengall bertbimodalselfsupervisedlearningforcropclassificationusingsentinel2andplanetscope AT juergengall bertbimodalselfsupervisedlearningforcropclassificationusingsentinel2andplanetscope |