What is it about?
The summary highlights a boundary-saving strategy designed to address memory usage challenges in automatic differentiation for seismic inversion. This approach effectively minimizes memory overhead, ensuring the accurate reconstruction of wavefields and precise gradient computation.
Featured Image
Why is it important?
This method reduces the memory consumption in FWI under the automatic differentiation framework to a level comparable to conventional implementations, enabling automatic differentiation to be applied to larger-scale models.
Perspectives
I hope to approach automatic differentiation technology from an integrated perspective, recognizing that some conventional techniques remain applicable within this framework.
Shaowen Wang
Ocean University of China
Read the Original
This page is a summary of: Memory Optimization in RNN-Based Full Waveform Inversion Using Boundary Saving Wavefield Reconstruction, IEEE Transactions on Geoscience and Remote Sensing, January 2023, Institute of Electrical & Electronics Engineers (IEEE),
DOI: 10.1109/tgrs.2023.3317529.
You can read the full text:
Contributors
The following have contributed to this page
