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

2025/10/08 by Karim El Khoury, Maxime Zanella, Khoury, Karim El +5 · 1 citation
Biochemistry, Genetics and Molecular Biology · Computer Science · Medicine · #Cancer-related molecular mechanisms research #Cervical Cancer and HPV Research #Computer Vision and Pattern Recognition (cs.CV) #Domain Adaptation and Few-Shot Learning #FOS: Computer and information sciences

paper · pdf · doi:10.48550/arxiv.2510.07135

openalex publication_date 2025/10/08 · openalex created_date 2025/10/18 · openalex updated_date 2026/07/28

Abstract

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

Citations

• Vocabulary-free few-shot learning for Vision-Language Models
• Enhancing Remote Sensing Vision-Language Models for Zero-Shot Scene Classification
• Low-Rank Few-Shot Adaptation of Vision-Language Models
• Streamlined Hybrid Annotation Framework using Scalable Codestream for Bandwidth-Restricted UAV Object Detection
• LHRS-Bot: Empowering Remote Sensing with VGI-Enhanced Large Multimodal Language Model
• A Closer Look at the Few-Shot Adaptation of Large Vision-Language Models
• SkyScript: A Large and Semantically Diverse Vision-Language Dataset for Remote Sensing
• Self-regulating Prompts: Foundational Model Adaptation without Forgetting
• RemoteCLIP: A Vision Language Foundation Model for Remote Sensing
• Visual-Language Prompt Tuning with Knowledge-guided Context Optimization
• Task Residual for Tuning Vision-Language Models
• MaPLe: Multi-modal Prompt Learning
• Tip-Adapter: Training-free Adaption of CLIP for Few-shot Classification
• Prompt-aligned Gradient for Prompt Tuning
• Conditional Prompt Learning for Vision-Language Models
• CLIP-Adapter: Better Vision-Language Models with Feature Adapters
• CLIP-Adapter: Better Vision-Language Models with Feature Adapters
• Learning to Prompt for Vision-Language Models
• Learning Transferable Visual Models From Natural Language Supervision
• MLRSNet: A Multi-label High Spatial Resolution Remote Sensing Dataset for Semantic Scene Understanding
• Deep learning in environmental remote sensing: Achievements and challenges
• AID: A Benchmark Dataset for Performance Evaluation of Aerial Scene Classification
• VHM: Versatile and Honest Vision Language Model for Remote Sensing Image Analysis

Cited by

• Locally Consistent Transductive Information Maximization for Few-Shot Remote Sensing Scene Classification

Related

• Enhancing Remote Sensing Vision-Language Models for Zero-Shot Scene Classification
• Efficient Few-Shot Learning in Remote Sensing: Fusing Vision and Vision-Language Models
• Few-Shot Remote Sensing Image Scene Classification with CLIP and Prompt Learning
• RemoteCLIP: A Vision Language Foundation Model for Remote Sensing
• Generalized Few-Shot Semantic Segmentation in Remote Sensing: Challenge and Benchmark