Symmetric autoencoders for retrieving void-generated ringing modes and backscatter from vehicle seismic noise

ABSTRACT Detecting deep voids in urban areas remains a challenging problem for geophysical methods, particularly with conventional active seismic sources that lack the low-frequency energy necessary for greater depth penetration. To address this, we explore an alternative approach that uses seismic noise generated by the engine of an idle vehicle to supply the needed low frequencies and improve the detection capabilities for voids at depths beyond 8 m. A key obstacle in processing this seismic noise is the presence of crosscorrelation residuals, which occur when traditional linear stacking methods struggle to achieve stable averages, particularly with data of shorter-than-usual time duration. These residuals can interfere with the identification of void signatures related to backscattering and ringing, making void detection challenging in dynamic urban settings. We use a symmetric autoencoder (SymAE), which is an unsupervised machine-learning architecture, to identify void signatures. By training neural networks to disentangle coherent subsurface information from the vehicle seismic noise, SymAE effectively minimizes the effects of crosscorrelation residuals and enhances the extraction of backscatters and ringing patterns. By applying this technique in a semicontrolled field experiment, we successfully detect two known voids at a depth of approximately 10 m. We further validate this approach using realistic synthetic data generated from a recorded vehicle signature. SymAE reduces data requirements while improving detection accuracy, showing the potential for rapid, cost effective, and reliable subsurface void detection in urban environments.