Cewen Liu
Cewen Liu｜ 刘策文
PhD candidate in
Department of Earth System Science
Tsinghua University
Email: Liucw17@mails.tsinghua.edu.cn
Address: Room S814, Mengminwei Building, Tsinghua University, Beijing, China
Education Background
Sep 2017 – Ongoing
Tsinghua University, Ph.D. candidate
Sep 2013 – June 2017
China University of Geosciences, B.S. in Geophysics
Intern Experience
Jul 2018 – Ongoing
BGP Beijing R&D center, Research Engineer
Jul 2017 – Sep 2017
National Supercomputer center in Wuxi, Research Engineer
Research Interest
Machine Learning in Geophysics
High resolution imaging, Point Spread Function(PSF), Model Inversion, CNN, ResNet
- Deep learning-based point spread function deconvolution for migration image deblurring. (C. Liu, M. Sun, et al. 2020, Geophysics Article, Moderate Revision)
- Enhance 3D seismic images resolution by deconvolving point spread function. (C. Liu and H Fu, First International Meeting for Applied Geoscience & Energy. 2021)
- Enhancing the Resolution of Seismic Imaging by Deconvolving Point Spread Function. (C. Liu, N. Dai et al. 2020, EAGE Annual conference)
- Inverting Elastic Model Properties Using ResNet. (C. Liu, M. Sun, et al. 2020, EAGE Digital Conference)
- see more from Geophysics
Full Waveform Inversion
Multi-scale, Source encoded, Cycle Skipping, Marmousi model
- We study the influence of different iterative methods on the results of full waveform inversion, and the influence of initial model, forward modeling and data noise on the accuracy of full waveform inversion. (Bachelor’s degree thesis, 2017)
- A source-encoded Full Waveform Inversion on BP model is tested, where multiscale strategy is used.
High Performance Geo-Computing
SW26010, Large Scale Earthquake; CUDA, Athread, MPI
- Large Scale Earthquake Simulation based on the SW-Taihulight Supercomputer
- CPC Competition based on the SW-Taihulight Supercomputer (September, 2020)
Competition & Awards
Secondary Scholarship, Tsinghua University, 2020
Bronze Award, The 4th domestic CPU parallel application challenge, 2020