Few-Shot Adaptation Benchmark for Remote Sensing Vision-Language Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Khoury, Karim El, Zanella, Maxime, De Vleeschouwer, Christophe, Macq, Benoit
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913173222719488
author Khoury, Karim El
Zanella, Maxime
De Vleeschouwer, Christophe
Macq, Benoit
author_facet Khoury, Karim El
Zanella, Maxime
De Vleeschouwer, Christophe
Macq, Benoit
contents Remote Sensing Vision-Language Models (RSVLMs) have shown remarkable potential thanks to large-scale pretraining, achieving strong zero-shot performance on various tasks. However, their ability to generalize in low-data regimes, such as few-shot learning, remains insufficiently explored. In this work, we present the first structured benchmark for evaluating few-shot adaptation methods on RSVLMs. We conduct comprehensive experiments across ten remote sensing scene classification datasets, applying five widely used few-shot adaptation strategies to three state-of-the-art RSVLMs with varying backbones. Our findings reveal that models with similar zero-shot performance can exhibit markedly different behavior under few-shot adaptation, with some RSVLMs being inherently more amenable to such adaptation than others. The variability of performance and the absence of a clear winner among existing methods highlight the need for the development of more robust methods for few-shot adaptation tailored to RS. To facilitate future research, we provide a reproducible benchmarking framework and open-source code to systematically evaluate RSVLMs under few-shot conditions. The source code is publicly available on Github: https://github.com/elkhouryk/fewshot_RSVLMs
format Preprint
id arxiv_https___arxiv_org_abs_2510_07135
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Few-Shot Adaptation Benchmark for Remote Sensing Vision-Language Models
Khoury, Karim El
Zanella, Maxime
De Vleeschouwer, Christophe
Macq, Benoit
Computer Vision and Pattern Recognition
Remote Sensing Vision-Language Models (RSVLMs) have shown remarkable potential thanks to large-scale pretraining, achieving strong zero-shot performance on various tasks. However, their ability to generalize in low-data regimes, such as few-shot learning, remains insufficiently explored. In this work, we present the first structured benchmark for evaluating few-shot adaptation methods on RSVLMs. We conduct comprehensive experiments across ten remote sensing scene classification datasets, applying five widely used few-shot adaptation strategies to three state-of-the-art RSVLMs with varying backbones. Our findings reveal that models with similar zero-shot performance can exhibit markedly different behavior under few-shot adaptation, with some RSVLMs being inherently more amenable to such adaptation than others. The variability of performance and the absence of a clear winner among existing methods highlight the need for the development of more robust methods for few-shot adaptation tailored to RS. To facilitate future research, we provide a reproducible benchmarking framework and open-source code to systematically evaluate RSVLMs under few-shot conditions. The source code is publicly available on Github: https://github.com/elkhouryk/fewshot_RSVLMs
title Few-Shot Adaptation Benchmark for Remote Sensing Vision-Language Models
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2510.07135