Numerical and experimental generated data during project https://doi.org/10.1038/s41598-024-65996-0

The dataset was generated using a deep-learning-based surrogate modeling technique for characterizing buried objects using 3-D full-wave electromagnetic simulations of a GPR model. The task was to independently predict characteristic parameters of a buried object of diverse radii allocated at different positions (depth and lateral position) in various dispersive subsurface media. The proposed surrogate model referred to as the deep regression network (DRN) is utilized for the time-frequency spectrogram (TFS) of consecutive A-scans. DRN is developed with the main aim being computationally efficient (about 13 times acceleration) compared to conventional network models using B-scan images (2D data). DRN with TFS is favorably benchmarked to the state-of-the-art regression techniques. The experimental results obtained for the proposed model and second-best model, CNN-1D show mean absolute and relative error rates of 3.6 mm, 11.8 mm, and 4.7%, 11.6%, respectively.