Reservoir-enhanced Segment Anything Model for Subsurface Diagnosis

2025-04-26

AI Summary

Urban subsurface anomaly detection (e.g., voids, cracks) using ground-penetrating radar (GPR) suffers from low accuracy and poor generalization due to label scarcity, ambiguous anomaly boundaries, and heterogeneous urban scenarios. To address these challenges, this paper proposes a few-shot segmentation framework tailored for GPR data. We innovatively integrate electromagnetic (EM) wave temporal/spatial variation modeling into the Segment Anything Model (SAM), introducing a “reservoir-enhanced” mechanism and a local EM-wave-variation attention module. The framework synergistically combines prompt learning, time-frequency domain waveform modeling, and lightweight feature distillation to enable end-to-end anomaly segmentation and fine-grained classification—without requiring target-domain annotations. Experiments demonstrate a detection accuracy of 85.3%, surpassing state-of-the-art methods; training data requirements are reduced by 70%; and the framework supports real-time inference and cross-scenario generalization.

Technology Category

• Computer Vision: Segmentation
• Data Mining & Knowledge Management: Anomaly/Outlier Detection
• Machine Learning: Multimodal Learning

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Urban roads and infrastructure, vital to city operations, face growing threats from subsurface anomalies like cracks and cavities. Ground Penetrating Radar (GPR) effectively visualizes underground conditions employing electromagnetic (EM) waves; however, accurate anomaly detection via GPR remains challenging due to limited labeled data, varying subsurface conditions, and indistinct target boundaries. Although visually image-like, GPR data fundamentally represent EM waves, with variations within and between waves critical for identifying anomalies. Addressing these, we propose the Reservoir-enhanced Segment Anything Model (Res-SAM), an innovative framework exploiting both visual discernibility and wave-changing properties of GPR data. Res-SAM initially identifies apparent candidate anomaly regions given minimal prompts, and further refines them by analyzing anomaly-induced changing information within and between EM waves in local GPR data, enabling precise and complete anomaly region extraction and category determination. Real-world experiments demonstrate that Res-SAM achieves high detection accuracy (>85%) and outperforms state-of-the-art. Notably, Res-SAM requires only minimal accessible non-target data, avoids intensive training, and incorporates simple human interaction to enhance reliability. Our research provides a scalable, resource-efficient solution for rapid subsurface anomaly detection across diverse environments, improving urban safety monitoring while reducing manual effort and computational cost.

Problem

• Detecting subsurface anomalies in urban infrastructure using GPR data
• Overcoming challenges of limited labeled data and unclear target boundaries
• Improving accuracy and efficiency in subsurface anomaly identification

Innovation

• Reservoir-enhanced SAM combines visual and wave data
• Minimal prompts refine anomaly detection via EM waves
• Resource-efficient, high-accuracy subsurface anomaly detection