Poster + Presentation
10 October 2020 Statistical analysis in machine learning reconstruction of images transmitted through a multimode fiber
Wenjian Lu
Author Affiliations +
Conference Poster
Abstract
We applied a one-dense-layer model to reconstruct the input images from the intensity of the speckle patterns coupled out from a multimode fiber. We studied the singular value statistics of the complex weight matrix, an approximation to the full transmission matrix. We compared the reconstruction quality with other DL models. We observed less time consuming and better tolerance to speckle patterns blocking and downsampling compared to other DL methods. We believe this one-dense-layer model shows alternative routes to retrieve the transmission matrix with machine learning. Our statistical study of the complex weight matrix would advance the physical understanding of image reconstruction through multimode fiber with different DL models.
Conference Presentation
© (2020) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Wenjian Lu "Statistical analysis in machine learning reconstruction of images transmitted through a multimode fiber", Proc. SPIE 11549, Advanced Optical Imaging Technologies III, 115491K (10 October 2020); https://doi.org/10.1117/12.2575177
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Machine learning

Multimode fibers

Speckle pattern

Statistical analysis

Image restoration

Inverse optics

Inverse problems

Back to Top