Communications of the ACM
Optical Computing for Object Classification Through Diffusive Random Media
The single-pixel broadband diffractive neural network classifies handwritten digits through unknown random diffusers.
Credit: Ozcan Lab/UCLA
University of California, Los Angeles (UCLA) researchers have developed an all-optical object classification technique that uses diffractive deep neural networks and a single-pixel spectral detector to directly classify unknown objects through unknown random diffusers.
The broadband diffractive network architecture developed by the researchers employs 20 discrete wavelengths to map a diffuser-distorted object into a spectral signature detected through a single pixel.
The researchers determined the illumination wavelength can be operated at any part of the electromagnetic spectrum without redesigning or retraining the diffractive layers.
Said UCLA's Aydogan Ozcan, "This work constitutes the first demonstration of all-optical classification of objects through random diffusers that generalizes to new unknown diffusers."
From UCLA Samueli School of Engineering
View Full Article
Abstracts Copyright © 2023 SmithBucklin, Washington, DC, USA
No entries found