Abstract
We present a one-time error-aware software training technique that outputs matrices that produce near-ideal performance on any arbitrary highly faulty photonic neural network with hardware phase errors up to 5x larger than present-day fabrication tolerances.
© 2022 The Author(s)
PDF Article | Presentation VideoMore Like This
James Spall, Xianxin Guo, and A. I. Lvovsky
FTu6D.2 Frontiers in Optics (FiO) 2022
Haoqin Deng and Mercedeh Khajavikhan
FM5B.5 CLEO: QELS_Fundamental Science (CLEO:FS) 2022
Zibo Hu, Russell L.T. Schwartz, Maria Solyanik-Gorgone, and Volker J Sorger
JW2A.18 3D Image Acquisition and Display: Technology, Perception and Applications (3D) 2022