INRA-Transformer: A Multi-Scale Hybrid Vision Transformer for Precise Seismic Deformation Mapping of the 6.1 Mw Sındırgı Earthquake
Abstract
Abstract Precise mapping of surface deformation is essential for seismic hazard assessment and tectonic process interpretation. Differential Interferometric Synthetic Aperture Radar (DInSAR) is a established geodetic technique for detecting ground displacement; however, conventional classifiers and standard deep learning models often struggle to resolve subtle deformation signals from complex phase noise and decorrelation artifacts. This study introduces INRA-Transformer, a novel hybrid Vision Transformer architecture designed to enhance post-seismic deformation mapping by integrating multi-scale feature extraction with global contextual reasoning. Focusing on the 6.1 Mw Sındırgı earthquake (Balıkesir, Türkiye), Sentinel-1A DInSAR interferograms were processed to evaluate model performance in a real-world seismic scenario. Unlike standard ViTs, the INRA-Transformer incorporates an Inception-based multi-scale patch embedding stem and a depthwise-convolutional feed-forward network (DW-FFN) to inject local inductive bias into the Transformer framework. Experimental results demonstrate that the proposed architecture achieves an overall accuracy (OA) of 97.58% and a Kappa coefficient of 0.966 across 50,854 validation samples. Class-wise evaluation highlights exceptional performance in identifying high-intensity deformation zones (F1-score} = 0.99). Furthermore, Explainable AI (XAI) attention mapping confirms that the network effectively suppresses decorrelation noise and…
Read the paper
Related studies
- Current state and future directions for deep learning based automatic seismic fault interpretation: A systematic review
- Physics of Partially Saturated Porous Media: Residual Saturation and Seismic-Wave Propagation
- Modern Imaging Using Seismic Reflection Data
- Effects of marine seismic surveys on free-ranging fauna: a systematic literature review
- A review of the recent seismic activity in the southern part of egypt (upper egypt)
- A review of research on mechanism of seismic attenuation and energy compensation
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.