RSGPNet: Geometric Prompting for Remote Sensing Open-Vocabulary Semantic Segmentation

2026-06-24

AI Summary

This work addresses the challenges of unstable local-region segmentation and poor handling of unseen categories in open-vocabulary semantic segmentation of remote sensing imagery by proposing a training-free geometric prompting framework. It introduces, for the first time, a geometric prompting mechanism into open-vocabulary remote sensing segmentation, leveraging CLIP-guided text prompts to generate coarse masks, which are subsequently refined through geometric region prompts. A coarse-to-fine consistency verification strategy is further incorporated to ensure segmentation accuracy across scales. Notably, the entire pipeline operates without any model fine-tuning and achieves state-of-the-art performance across multiple remote sensing benchmarks, demonstrating superior stability and interpretability—particularly in regions with complex boundaries.

Abstract

Open-vocabulary semantic segmentation (OVSS) enables text-guided segmentation of unseen objects, breaking fixed-class limitations to achieve open-world understanding. However, existing OVSS methods primarily focus on modifying the CLIP attention mechanism, which still suffers from unstable local segmentation for remote sensing (RS) domain. To address these limitations, we propose RSGPNet, a training-free geometric prompting framework for RS OVSS that refines segmentation by leveraging object geometric areas and consistency constraints. Specifically, RSGPNet comprises three core modules: a Text-guided Coarse Mask module (TCM), a Geometric Re-prompting Module (GRP), and a Coarse-to-fine Consistency Verification Mechanism (CVM). TCM utilizes text prompts and the input image to construct initial coarse segmentation masks. GRP then converts these coarse masks into geometric box prompts, feeding them back into the segmentation model to generate refined masks. Finally, CVM employs consistency computation to prevent prompting from reinforcing erroneous regions. They allow the model to improve segmentation accuracy in complex areas, such as category boundaries. Extensive experiments on RS datasets demonstrate that RSGPNet significantly outperforms state-of-the-art methods across both quantitative and qualitative metrics while exhibiting excellent interpretability. The code is released at \href{https://github.com/wangshanwen001/RSGPNet}{https://github.com/wangshanwen001/RSGPNet}.

Problem

• open-vocabulary semantic segmentation
• remote sensing
• unstable local segmentation
• geometric prompting
• CLIP attention mechanism

Innovation

• geometric prompting
• open-vocabulary semantic segmentation
• remote sensing
• training-free framework
• consistency verification