MetaMamba: Meta-learning with mamba for few-shot vegetation species classification using UAV-based hyperspectral imagery in desert rangeland
Fei Hao, Xinchao Gao, Tao Zhang, Jianmin Du, Shengli Wang, Xiaoyan He · 2026
Read MetaMamba: Meta-learning with mamba for few-shot vegetation species classification using UAV-based hyperspectral imagery in desert rangeland by Fei Hao, Xinchao Gao, Tao Zhang, Jianmin Du, Shengli Wang, Xiaoyan He (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/plos-pone-0352744.txt.
Description
Fei Hao and co-authors build a drone-based system for naming desert grassland plants from the air when only a handful of labeled examples exist. Desert rangelands need watching for ecological security, and hyperspectral cameras see species-level detail, but training data are scarce. Their method learns how to learn. One branch reads fine local texture. A second, sequence-based branch reads wide spatial patterns. A meta-learning routine then teaches the pair to adapt from few examples.
Across test scenes the system classified species with accuracy above ninety percent on headline measures and stayed steady as sample sizes shrank. Hao and colleagues present it as a practical route for agencies that must map fragile rangelands without funding huge labeling campaigns first.
Catalog record
| CREATOR | Fei Hao, Xinchao Gao, Tao Zhang, Jianmin Du, Shengli Wang, Xiaoyan He |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 53925 |
| LICENSE | CC_BY |
| RIGHTS | CC BY |
| SOURCE | PLOS |