Deep Convolutional Neural Network Approach for Solving Nonlinear Inverse Scattering Problems
- Resource Type
- Authors
- Lianlin Li; Fernando L. Teixeira; Long Gang Wang; Daniel Ospina Acero
- Source
- 2019 IEEE International Symposium on Antennas and Propagation and USNC-URSI Radio Science Meeting.
- Subject
- business.industry
Computer science
Deep learning
Computer Science::Neural and Evolutionary Computation
Structure (category theory)
02 engineering and technology
Function (mathematics)
Similarity measure
01 natural sciences
Convolutional neural network
010101 applied mathematics
Nonlinear system
Inverse scattering problem
0202 electrical engineering, electronic engineering, information engineering
020201 artificial intelligence & image processing
Artificial intelligence
0101 mathematics
business
Algorithm
- Language
We describe a deep convolutional neural network (CNN) approach for nonlinear electromagnetic (EM) inverse scattering problems. We evaluate the performance of the proposed CNN as a function of the number of layers using different metrics such as structure similarity measure and mean-square error. The results show that the proposed DNN architecture has great potential in tackling nonlinear inverse scattering problems.