Deep learning–based vortex decomposition and switching based on fiber vector eigenmodes
Hou Mengdie,
Xu Mengjun,
Xu Jiangtao,
Lu Jiafeng,
An Yi,
Huang Liangjin,
Zeng Xianglong,
Pang Fufei,
Li Jun,
Yi Lilin
Affiliations
Hou Mengdie
The Key Lab of Specialty Fiber Optics and Optical Access Network, Joint International Research Laboratory of Specialty Fiber Optics and Advanced Communication, Shanghai University, Shanghai200444, China
Xu Mengjun
The Key Lab of Specialty Fiber Optics and Optical Access Network, Joint International Research Laboratory of Specialty Fiber Optics and Advanced Communication, Shanghai University, Shanghai200444, China
Xu Jiangtao
The Key Lab of Specialty Fiber Optics and Optical Access Network, Joint International Research Laboratory of Specialty Fiber Optics and Advanced Communication, Shanghai University, Shanghai200444, China
Lu Jiafeng
The Key Lab of Specialty Fiber Optics and Optical Access Network, Joint International Research Laboratory of Specialty Fiber Optics and Advanced Communication, Shanghai University, Shanghai200444, China
An Yi
College of Advanced Interdisciplinary Studies, National University of Defense Technology, Changsha410073, China
Huang Liangjin
Nanhu Laser Laboratory, College of Advanced Interdisciplinary Studies, National University of Defense Technology, Changsha410073, China
Zeng Xianglong
The Key Lab of Specialty Fiber Optics and Optical Access Network, Joint International Research Laboratory of Specialty Fiber Optics and Advanced Communication, Shanghai University, Shanghai200444, China
Pang Fufei
The Key Lab of Specialty Fiber Optics and Optical Access Network, Joint International Research Laboratory of Specialty Fiber Optics and Advanced Communication, Shanghai University, Shanghai200444, China
Li Jun
College of Advanced Interdisciplinary Studies, National University of Defense Technology, Changsha410073, China
Yi Lilin
State Key Laboratory of Advanced Optical Communication Systems and Networks, Shanghai Jiao Tong University, Shanghai200240, China
Structured optical fields, such as cylindrical vector (CV) and orbital angular momentum (OAM) modes, have attracted considerable attention due to their polarization singularities and helical phase wavefront structure. However, one of the most critical challenges is still the intelligent generation or precise control of these modes. Here, we demonstrate the first simulation and experimental realization of decomposing the CV and OAM modes by reconstructing the multi-view images of projected intensity distribution. Assisted by the deep learning–based stochastic parallel gradient descent (SPGD) algorithm, the modal coefficients and optical field distributions can be retrieved in 1.32 s within an average error of 0.416 % showing high efficiency and accuracy. Especially, the interference pattern and quarter-wave plate are exploited to confirm the phase and distinguish elliptical or circular polarization direction, respectively. The generated donut modes are experimentally decomposed in the CV and OAM modes, where purity of CV modes reaches 99.5 %. Finally, fast switching vortex modes is achieved by electrically driving the polarization controller to deliver diverse CV modes. Our findings may provide a convenient way to characterize and deepen the understanding of CV or OAM modes in view of modal proportions, which is expected of latent applied value on information coding and quantum computation.