Abstract
We demonstrate an adaptive learning framework, termed dynamic synthesis network (DSN), which dynamically synthesizes model weights and adapts to different scattering conditions. The efficiency of DSN is demonstrated in holographic 3D particle descattering.
© 2022 The Author(s)
PDF ArticleMore Like This
Myung K. Kim
SM3F.1 Optical Sensors (Sensors) 2012
Yuhong Wan, Hongqiang Zhou, and Tianlong Man
W4G.3 Conference on Lasers and Electro-Optics/Pacific Rim (CLEO/PR) 2018
Chencen Xiong, Zhenbo Ren, Jianglei Di, and Jianlin Zhao
W5A.52 Digital Holography and Three-Dimensional Imaging (DH) 2022