# SG-DeepONet: Source-generalized deep operator learning for full waveform inversion Source: [SG-DeepONet: Source-generalized deep operator learning for full waveform inversion]() Collection index: [llms.txt]() Data updated: 2026-03-19T05:00:05.030Z Authors: Zekai Guo, Lihui Chai, Ye Li Published: 2024-08-15T08:15:06.000Z ## Abstract Full waveform inversion (FWI) aims to reconstruct subsurface velocity models from observed seismic wavefields and has recently benefited from advances in deep learning (DL). The performance of DL-based FWI critically depends on the diversity of training data, yet existing datasets such as OpenFWI rely on fixed or weakly varying source conditions, limiting their ability to represent realistic seismic scenarios and hindering source generalization. To address this issue, we construct a new source-variable seismic dataset, termed SVFWI, by systematically varying the frequencies and horizontal locations of multiple surface sources. SVFWI is further divided into three subsets that respectively model frequency variations, location variations, and their combined effects, providing a challenging benchmark in data-driven FWI. We further propose SG-DeepONet, a novel DeepONet-based encoder-decoder framework tailored for FWI. The branch network extracts multi-scale time-frequency features from seismic observations, the trunk network explicitly embeds source physical parameters, and an interactive decoding network enables effective nonlinear fusion and high-fidelity velocity reconstruction. Extensive experiments on SVFWI demonstrate that SG-DeepONet achieves superior inversion accuracy and robustness under varying source conditions compared with existing DL-based FWI methods. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Full Waveform Inversion on CurveVel-A SVFWI-FL (test)](): Full Waveform Inversion; CurveVel-A SVFWI-FL (test); MAE: 0.058. - [Full Waveform Inversion on FlatVel-B SVFWI-L](): Full Waveform Inversion; FlatVel-B SVFWI-L; MAE: 2.56. - [Full Waveform Inversion on CurveVel-A SVFWI-L](): Full Waveform Inversion; CurveVel-A SVFWI-L; MAE: 0.0374. - [Full Waveform Inversion on FlatFault-B (SVFWI-L)](): Full Waveform Inversion; FlatFault-B (SVFWI-L); MAE: 0.0754. - [Full Waveform Inversion on CurveFault-A SVFWI-L](): Full Waveform Inversion; CurveFault-A SVFWI-L; MAE: 0.0211. - [Full Waveform Inversion on SVFWI-F FlatVel-B (test)](): Full Waveform Inversion; SVFWI-F FlatVel-B (test); MAE: 0.0501. - [Full Waveform Inversion on SVFWI-F FlatFault-B (test)](): Full Waveform Inversion; SVFWI-F FlatFault-B (test); MAE: 0.0983. - [Full Waveform Inversion on SVFWI-F CurveFault-A (test)](): Full Waveform Inversion; SVFWI-F CurveFault-A (test); MAE: 2.66. ## Original sources and implementation [Original paper]() [PDF]()