A Nonlinear Model Compression Scheme Based on Variational Autoencoder for Microwave Data Inversion
- 8 August 2022
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Antennas and Propagation
- Vol. 70 (11) , 11059-11069
Abstract
We present an inversion algorithm with a deep-learning-based model compression scheme. Models are described with latent parameters of a trained variational autoencoder (VAE) neural network. Given observed data, latent parameters are inverted by minimizing the data misfit cost function using the Gauss-Newton method. This inversion algorithm is tested using both synthetic and experimental datasets. We achieve a 0.87% compression rate while maintaining high-quality reconstruction. The deep neural network renders nonlinear model compression, which largely reduces the number of unknowns; hence, it has higher computational efficiency. Furthermore, various prior knowledge that is difficult to describe with rigorous forms can be incorporated into inversion through training the neural network, which mitigates the ill-posedness of the inverse problem.Keywords
Funding Information
- National Key Research and Development Program of China (2018YFC0603604)
- National Natural Science Foundation of China (61971263)
- Shuimu Tsinghua Scholar Program of Tsinghua University (2021SM031)
- Beijing Innovation Center for Future Chip, Institute for Precision Medicine, Tsinghua University
- Biren Tech, Beijing, China
This publication has 43 references indexed in Scilit:
- A contrast source inversion method in the wavelet domainInverse Problems, 2013
- Application of a two-and-a-half dimensional model-based algorithm to crosswell electromagnetic data inversionInverse Problems, 2010
- 2.5D forward and inverse modeling for interpreting low-frequency electromagnetic measurementsGeophysics, 2008
- Level set methods for inverse scatteringInverse Problems, 2006
- A Microwave Imaging Method for NDE/NDT Based on the SMW Technique for the Electromagnetic Field PredictionIEEE Transactions on Instrumentation and Measurement, 2006
- A GENERAL FRAMEWORK FOR CONSTRAINT MINIMIZATION FOR THE INVERSION OF ELECTROMAGNETIC MEASUREMENTSProgress In Electromagnetics Research, 2004
- Imaging of biomedical data using a multiplicative regularized contrast source inversion methodIEEE Transactions on Microwave Theory and Techniques, 2002
- On the local minima in a tomographic imaging techniqueIEEE Transactions on Geoscience and Remote Sensing, 2001
- Contrast Source Inversion Method: State of ArtProgress In Electromagnetics Research, 2001
- Inverse scattering of dielectric cylinders by a second-order Born approximationIEEE Transactions on Geoscience and Remote Sensing, 1999