This work proposes a deep learning (DL)-based framework, namely Sim2Real, for spectral signal reconstruction in reconstructive spectroscopy, focusing on efficient data sampling and fast inference time. The work focuses on the challenge of reconstructing real-world spectral signals in an extreme setting where only device-informed simulated data are available for training. Such device-informed simulated data are much easier to collect than real-world data but exhibit large distribution shifts from their real-world counterparts. To leverage such simulated data effectively, a hierarchical data augmentation strategy is introduced to mitigate the adverse effects of this domain shift, and a corresponding neural network for the spectral signal reconstruction with our augmented data is designed. Experiments using a real dataset measured from our spectrometer device demonstrate that Sim2Real achieves significant speed-up during the inference while attaining on-par performance with the state-of-the-art optimization-based methods.
Note: When clicking on a Digital Object Identifier (DOI) number, you will be taken to an external site maintained by the publisher.
Some full text articles may not yet be available without a charge during the embargo (administrative interval).
What is a DOI Number?
Some links on this page may take you to non-federal websites. Their policies may differ from this site.
-
Free, publicly-accessible full text available September 1, 2025
-
Crocombe, Richard A ; Barnett, Steven M ; Profeta, Luisa_T M (Ed.)Free, publicly-accessible full text available June 7, 2025
-
A variable transmission thin film for visible light is proposed based on a mechanically actuated origami structure coated with metallic nanoparticles. The transmissivity can be tuned continuously from 0 to
for unpolarized incident light. Power is only required for switching and is not necessary to maintain the desired transmittance state. The asymmetric metal nanorods create two distinct plasmon resonances. Controlling the orientation of the nanorods with respect to the direction of the incident light changes the optical transmittance. The switching speed is only limited by the mechanical actuation and not by the optical response of the materials. The applicability of the proposed film for smart glass applications is investigated. Good image transmission clarity with minimal distortion is shown.