Full Waveform Inversion

Full waveform inversion (FWI) is a geophysical imaging technique aiming to reconstruct subsurface structures by matching observed and simulated seismic waveforms. Current research emphasizes improving FWI's accuracy and efficiency through deep learning, employing architectures like DeepONets, encoder-decoder networks, and generative models (e.g., diffusion models and transformers) to handle the inherent non-linearity and uncertainty of the inverse problem. These advancements are driven by the need for higher-resolution subsurface imaging in applications such as resource exploration, carbon capture, and medical imaging (e.g., ultrasound computed tomography), where faster and more robust inversion methods are crucial. The development of large-scale benchmark datasets is also a significant focus, enabling more rigorous evaluation and comparison of different FWI approaches.