---
title: "PLANTEUR: A Crop-Type Segmentation Benchmark for Tropical Regions"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2611.05223v1"
permalink: "https://feed.astralor.com/items/itm_a3eb23c727a47aa9480b811e"
published: "2026-11-03T17:10:42.000Z"
score: 54
channel: "tech"
tags: ["开源","PLANTEUR","Crop-Type Segmentation Benchmark","Tropical Regions","French","Building","PASTIS"]
---

# PLANTEUR: A Crop-Type Segmentation Benchmark for Tropical Regions

## At a glance

We introduce PLANTEUR, a benchmark for crop-type mapping from satellite images in the French overseas territories. Building on the PASTIS design, PLANTEUR provides georeferenced and aligned Sentinel-2, Sentinel-1, and L…

## Key points

- We introduce PLANTEUR, a benchmark for crop-type mapping from satellite images in the French overseas territories.
- Building on the PASTIS design, PLANTEUR provides georeferenced and aligned Sentinel-2, Sentinel-1, and Landsat-8 sequences with parcel-derived labels spanning six tropical regions and an 8-class label space.
- We define both a standard test split and a leave-one-island-out protocol to measure cross-territory generalization, and evaluate fully supervised models, linear probes, and fine-tuned foundation models.

## Full text

We introduce PLANTEUR, a benchmark for crop-type mapping from satellite images in the French overseas territories. Building on the PASTIS design, PLANTEUR provides georeferenced and aligned Sentinel-2, Sentinel-1, and Landsat-8 sequences with parcel-derived labels spanning six tropical regions and an 8-class label space. We define both a standard test split and a leave-one-island-out protocol to measure cross-territory generalization, and evaluate fully supervised models, linear probes, and fine-tuned foundation models. The benchmark code and data are open source and available on GitHub and Hugging Face.

---

Verify against the original before citing; curated by Astral Feed: https://feed.astralor.com/items/itm_a3eb23c727a47aa9480b811e