My research interest is mostly focused on seismic imaging and macro velocity model building. In Aramco, I am part of the project for 3D advanced Reverse Time Migration, I developed a code for compressing the wavefield and the compression ratio can be up to 40x. The code is routinely used in production by Aramco, and Aramco are applying for a US patent for this efficient compression technology:
Compressing seismic wavefields in three-dimensional reverse time migration. Saudi Aramco, first author.
In KAUST, I developed a Robust method of combining optimal transport and matching filter in full-waveform inversion for reducing cycle skipping. This technique is applying for a US patent currently:
A method for a robust full waveform inversion. KAUST, first author.
Now, as AI becomes popular, I am also interested in applying machine learning technology to seismic imaging and inversion problem. Specially, I designed to using neural network to formulate a misfit function in full waveform inversion and such technology is applying US patent:
A method for using a neural network to formulate an optimization problem. KAUST, first author.
Neural networks were also adjusted to formulate a generate framework for processing seismic datasets:
A framework for seismic data processing using machine learning. KAUST, first author.