Fiber-based diffractive deep neural network
Optics Letters · 01.09.2025 · 50 (17) · 5254
DOI: 10.1364/ol.570135 ↗Kurz gefasst
Fachartikel in der Zeitschrift Optics Letters, online erschienen am 19.08.2025. Verfasst von Bahadır Utku Kesgin und 2 weiteren Autor:innen. Die Arbeit wurde bisher 11-mal zitiert (Crossref) und nennt 34 Literaturangaben. Gefördert von: Türkiye Bilimsel ve Teknolojik Araştırma Kurumu.
Abstract (Originalsprache)
Optical computing has reemerged as a promising alternative computing paradigm for providing energy-efficient information processing in the age of artificial intelligence. Among various photonic neural network platforms, diffractive optical processing systems in free space proved to be high-performance computing with high parallelism. Here, we report fiber-based diffractive deep neural networks by optimizing the linear coupling of the waveguide modes. Our approach demonstrated high performance in various machine learning tasks, such as biomedical disease, fashion, and geospatial classification, with a simple readout layer and all-optically. Operating on linear optics, our architecture performs on par with neural networks even in complex datasets where the data cannot be separated using linear operations. These results will enable efficient and scalable diffractive information processing with waveguides for real-life computing, telecommunications, and imaging applications.
Förderung
- Türkiye Bilimsel ve Teknolojik Araştırma Kurumu
Zitieren
APA
Kesgin, B. U., Yüce, F., & Teğin, U. (2025). Fiber-based diffractive deep neural network. Optics Letters, 50(17), 5254. https://doi.org/10.1364/ol.570135
Vancouver
Kesgin BU, Yüce F, Teğin U. Fiber-based diffractive deep neural network. Optics Letters. 2025;50(17):5254. doi:10.1364/ol.570135
BibTeX
@article{Kesgin2025,
author = {Kesgin, Bahadır Utku and Yüce, Firdevs and Teğin, Uğur},
title = {Fiber-based diffractive deep neural network},
journal = {Optics Letters},
year = {2025},
volume = {50},
number = {17},
pages = {5254},
doi = {10.1364/ol.570135}
}
EUVIMED