arXiv：cs.LG· Gedeon Muhawenayo, Caleb Robinson, Subash Khanal, Zhanpei Fang, Isaac Corley, Alexander Wollam, Tianyi Gao, Leonard Strnad, Ryan Avery, Lyndon Estes, Ana M. T\'arano, Nathan Jacobs, Hannah Kerner·· 4 小时前AI 评分37
PRUE：面向大规模田块边界分割的实用方案
PRUE: A Practical Recipe for Field Boundary Segmentation at Scale
AI 导读
PRUE 结合 U-Net 主干、复合损失函数与针对性数据增强，在 Fields of The World（FTW）基准上达到 76% IoU 和 47% object-F1，较此前基线分别提升 6% 和 9%。
正文
Authors:Gedeon Muhawenayo, Caleb Robinson, Subash Khanal, Zhanpei Fang, Isaac Corley, Alexander Wollam, Tianyi Gao, Leonard Strnad, Ryan Avery, Lyndon Estes, Ana M. Tárano, Nathan Jacobs, Hannah Kerner
Abstract:Large-scale maps of field boundaries are essential for agricultural monitoring tasks. Existing deep learning approaches for satellite-based field mapping are sensitive to illumination, spatial scale, and changes in geographic location. We conduct the first systematic evaluation of segmentation and geospatial foundation models (GFMs) for global field boundary delineation using the Fields of The World (FTW) benchmark. We evaluate 18 models under unified experimental settings, showing that a U-Net semantic segmentation model outperforms instance-based and GFM alternatives on a suite of performance and deployment metrics. We propose a new segmentation approach that combines a U-Net backbone, composite loss functions, and targeted data augmentations to enhance performance and robustness under real-world conditions. Our model achieves a 76% IoU and 47% object-F1 on FTW, an increase of 6% and 9% over the previous baseline. Our approach provides a practical framework for reliable, scalable, and reproducible field boundary delineation across model design, training, and inference. We release all models and model-derived field boundary datasets for five countries.
| Comments: | 12 pages, 3 figures, supplementary material. Accepted at CVPR 2026 (IEEE/CVF Conference on Computer Vision and Pattern Recognition) |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG) |
| Cite as: | arXiv:2603.27101 [cs.CV] |
| (or arXiv:2603.27101v2 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2603.27101 arXiv-issued DOI via DataCite |
|
| Journal reference: | Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026, pp. 6484-6495 |
Submission history
From: Gedeon Muhawenayo Mr. [view email]
[v1]
Sat, 28 Mar 2026 02:47:46 UTC (5,402 KB)
[v2]
Tue, 6 Oct 2026 17:17:19 UTC (5,402 KB)
来源：arXiv：cs.LG · arxiv.org