Dynamic gating-enhanced deep learning model with multi-source remote sensing synergy for optimizing wheat yield estimation
STF-MoE effectively handles multi-source spatiotemporal complexity via its dynamic gating and expert specialization. While underestimation persists in extreme-yield regions, the model provides a scalable solution for pre-harvest yield estimation. Future work will optimize computational efficiency and integrate higher-resolution data.
Authors
- Liu He (ORCID: https://orcid.org/0000-0002-2085-3970)
- Xue Lin (ORCID: https://orcid.org/0000-0001-7203-4929)
- Haibing Guan (ORCID: https://orcid.org/0000-0002-4714-7400)
- Jiawei Zhao (ORCID: https://orcid.org/0000-0001-7119-4082)
- Jian Lü (ORCID: https://orcid.org/0000-0002-9144-7486)
- Jian Li (ORCID: https://orcid.org/0000-0002-0864-5108)
- Zhihan Liu (ORCID: https://orcid.org/0000-0002-7449-1725)
- Zheng Li (ORCID: https://orcid.org/0000-0002-3078-1886)
- J. S. Kang (ORCID: https://orcid.org/0000-0002-5679-5338)
- Hongkun Fu
- B. Liu (ORCID: https://orcid.org/0000-0002-3006-5564)
Institutions
- Northeast Institute of Geography and Agroecology (CN)
- Jilin Agricultural University (CN)
Publication Details
- Journal
- Frontiers in Plant Science
- Published
- 2025-07-21
- DOI
- https://doi.org/10.3389/fpls.2025.1640806
- Citations
- 2
- Primary Topic
- Smart Agriculture and AI
- Type
- article
- Field-Weighted Citation Impact
- 3.50
Funders
- People's Government of Jilin Province
- Modern Agricultural Technology Industry System of Shandong province