We propose a design method for a diffractive neural network(DNN)for imaging through scattering media,offering robustness against the spatial coherence of illumination,scattering strength,and scattering dynamics.Most t...We propose a design method for a diffractive neural network(DNN)for imaging through scattering media,offering robustness against the spatial coherence of illumination,scattering strength,and scattering dynamics.Most techniques for imaging through scattering media are time-consuming and/or tailored to specific optical conditions.The DNN,composed of layers of diffractive optical elements(DOEs),optically reproduces the intensity distributions of objects behind scattering media without any computational processing.Datasets with randomized optical parameters are provided during the training process to achieve this robustness.We demonstrate the proposed method through numerical calculations and show its promising capability for DOE design.Our study paves the way for unifying and generalizing techniques for imaging through scattering media,which are currently fragmented by specific scenarios,enabling highly flexible imaging independent of optical conditions.展开更多
Optical computing is considered a promising solution for the growing demand for parallel computing in various cutting-edge fields that require high integration and high-speed computational capacity.We propose an optic...Optical computing is considered a promising solution for the growing demand for parallel computing in various cutting-edge fields that require high integration and high-speed computational capacity.We propose an optical computation architecture called diffraction casting(DC)for flexible and scalable parallel logic operations.In DC,a diffractive neural network is designed for single instruction,multiple data(SIMD)operations.This approach allows for the alteration of logic operations simply by changing the illumination patterns.Furthermore,it eliminates the need for encoding and decoding of the input and output,respectively,by introducing a buffer around the input area,facilitating end-to-end all-optical computing.We numerically demonstrate DC by performing all 16 logic operations on two arbitrary 256-bit parallel binary inputs.Additionally,we showcase several distinctive attributes inherent in DC,such as the benefit of cohesively designing the diffractive elements for SIMD logic operations that assure high scalability and high integration capability.Our study offers a design architecture for optical computers and paves the way for a next-generation optical computing paradigm.展开更多
基金Japan Society for the Promotion of Science(JP20H05890,JP22H05197,JP23H05444,JP23K26567)。
文摘We propose a design method for a diffractive neural network(DNN)for imaging through scattering media,offering robustness against the spatial coherence of illumination,scattering strength,and scattering dynamics.Most techniques for imaging through scattering media are time-consuming and/or tailored to specific optical conditions.The DNN,composed of layers of diffractive optical elements(DOEs),optically reproduces the intensity distributions of objects behind scattering media without any computational processing.Datasets with randomized optical parameters are provided during the training process to achieve this robustness.We demonstrate the proposed method through numerical calculations and show its promising capability for DOE design.Our study paves the way for unifying and generalizing techniques for imaging through scattering media,which are currently fragmented by specific scenarios,enabling highly flexible imaging independent of optical conditions.
基金supported by Japan Society for the Promotion of Science(Grant Nos.JP20K05361,JP22H05197,and JP23K26567).
文摘Optical computing is considered a promising solution for the growing demand for parallel computing in various cutting-edge fields that require high integration and high-speed computational capacity.We propose an optical computation architecture called diffraction casting(DC)for flexible and scalable parallel logic operations.In DC,a diffractive neural network is designed for single instruction,multiple data(SIMD)operations.This approach allows for the alteration of logic operations simply by changing the illumination patterns.Furthermore,it eliminates the need for encoding and decoding of the input and output,respectively,by introducing a buffer around the input area,facilitating end-to-end all-optical computing.We numerically demonstrate DC by performing all 16 logic operations on two arbitrary 256-bit parallel binary inputs.Additionally,we showcase several distinctive attributes inherent in DC,such as the benefit of cohesively designing the diffractive elements for SIMD logic operations that assure high scalability and high integration capability.Our study offers a design architecture for optical computers and paves the way for a next-generation optical computing paradigm.
