In-situ training in programmable photonic frequency circuits

Publikation: Beitrag in FachzeitschriftArtikelForschungPeer-Review

Autorschaft

  • Philip Rübeling
  • Oleksandr V. Marchukov
  • Filipe F. Bellotti
  • Ulrich B. Hoff
  • Nikolaj T. Zinner
  • Michael Kues
Forschungs-netzwerk anzeigen

Details

OriginalspracheEnglisch
Seiten (von - bis)2779-2786
Seitenumfang8
FachzeitschriftNanophotonics
Jahrgang14
Ausgabenummer16
Frühes Online-Datum23 Juni 2025
PublikationsstatusVeröffentlicht - 2 Aug. 2025

Abstract

Optical artificial neural networks (OANNs) leverage the advantages of photonic technologies including high processing speeds, low energy consumption, and mass production to establish a competitive and scalable platform for machine learning applications. While recent advancements have focused on harnessing spatial or temporal modes of light, the frequency domain attracts a lot of attention, with current implementations including spectral multiplexing, neural networks in nonlinear optical systems and extreme learning machines. Here, we present an experimental realization of a programmable photonic frequency circuit, realized with fiber-optical components, and implement the in-situ training with optical weight control of an OANN operating in the frequency domain. Input data is encoded into phases of frequency comb modes, and programmable phase and amplitude manipulations of the spectral modes enable in-situ training of the OANN, without employing a digital model of the device. The trained OANN achieves multiclass classification accuracies exceeding 90 %, comparable to conventional machine learning approaches. This proof-of-concept demonstrates the feasibility of a multilayer OANN in the frequency domain and can be extended to a scalable, integrated photonic platform with ultrafast weights updates, with potential applications to single-shot classification in spectroscopy.

ASJC Scopus Sachgebiete

Ziele für nachhaltige Entwicklung

Zitieren

In-situ training in programmable photonic frequency circuits. / Rübeling, Philip; Marchukov, Oleksandr V.; Bellotti, Filipe F. et al.
in: Nanophotonics, Jahrgang 14, Nr. 16, 02.08.2025, S. 2779-2786.

Publikation: Beitrag in FachzeitschriftArtikelForschungPeer-Review

Rübeling, P, Marchukov, OV, Bellotti, FF, Hoff, UB, Zinner, NT & Kues, M 2025, 'In-situ training in programmable photonic frequency circuits', Nanophotonics, Jg. 14, Nr. 16, S. 2779-2786. https://doi.org/10.1515/nanoph-2025-0125
Rübeling, P., Marchukov, O. V., Bellotti, F. F., Hoff, U. B., Zinner, N. T., & Kues, M. (2025). In-situ training in programmable photonic frequency circuits. Nanophotonics, 14(16), 2779-2786. https://doi.org/10.1515/nanoph-2025-0125
Rübeling P, Marchukov OV, Bellotti FF, Hoff UB, Zinner NT, Kues M. In-situ training in programmable photonic frequency circuits. Nanophotonics. 2025 Aug 2;14(16):2779-2786. Epub 2025 Jun 23. doi: 10.1515/nanoph-2025-0125
Rübeling, Philip ; Marchukov, Oleksandr V. ; Bellotti, Filipe F. et al. / In-situ training in programmable photonic frequency circuits. in: Nanophotonics. 2025 ; Jahrgang 14, Nr. 16. S. 2779-2786.
Download
@article{612f9f227348475582f345da6a7c1eb4,
title = "In-situ training in programmable photonic frequency circuits",
abstract = "Optical artificial neural networks (OANNs) leverage the advantages of photonic technologies including high processing speeds, low energy consumption, and mass production to establish a competitive and scalable platform for machine learning applications. While recent advancements have focused on harnessing spatial or temporal modes of light, the frequency domain attracts a lot of attention, with current implementations including spectral multiplexing, neural networks in nonlinear optical systems and extreme learning machines. Here, we present an experimental realization of a programmable photonic frequency circuit, realized with fiber-optical components, and implement the in-situ training with optical weight control of an OANN operating in the frequency domain. Input data is encoded into phases of frequency comb modes, and programmable phase and amplitude manipulations of the spectral modes enable in-situ training of the OANN, without employing a digital model of the device. The trained OANN achieves multiclass classification accuracies exceeding 90{\^a}€¯%, comparable to conventional machine learning approaches. This proof-of-concept demonstrates the feasibility of a multilayer OANN in the frequency domain and can be extended to a scalable, integrated photonic platform with ultrafast weights updates, with potential applications to single-shot classification in spectroscopy.",
keywords = "machine learning, photonic computing, ultrafast optics",
author = "Philip R{\"u}beling and Marchukov, {Oleksandr V.} and Bellotti, {Filipe F.} and Hoff, {Ulrich B.} and Zinner, {Nikolaj T.} and Michael Kues",
note = "Publisher Copyright: {\textcopyright} 2025 the author(s), published by De Gruyter, Berlin/Boston 2025.",
year = "2025",
month = aug,
day = "2",
doi = "10.1515/nanoph-2025-0125",
language = "English",
volume = "14",
pages = "2779--2786",
journal = "Nanophotonics",
issn = "2192-8606",
publisher = "Walter de Gruyter GmbH",
number = "16",

}

Download

TY - JOUR

T1 - In-situ training in programmable photonic frequency circuits

AU - Rübeling, Philip

AU - Marchukov, Oleksandr V.

AU - Bellotti, Filipe F.

AU - Hoff, Ulrich B.

AU - Zinner, Nikolaj T.

AU - Kues, Michael

N1 - Publisher Copyright: © 2025 the author(s), published by De Gruyter, Berlin/Boston 2025.

PY - 2025/8/2

Y1 - 2025/8/2

N2 - Optical artificial neural networks (OANNs) leverage the advantages of photonic technologies including high processing speeds, low energy consumption, and mass production to establish a competitive and scalable platform for machine learning applications. While recent advancements have focused on harnessing spatial or temporal modes of light, the frequency domain attracts a lot of attention, with current implementations including spectral multiplexing, neural networks in nonlinear optical systems and extreme learning machines. Here, we present an experimental realization of a programmable photonic frequency circuit, realized with fiber-optical components, and implement the in-situ training with optical weight control of an OANN operating in the frequency domain. Input data is encoded into phases of frequency comb modes, and programmable phase and amplitude manipulations of the spectral modes enable in-situ training of the OANN, without employing a digital model of the device. The trained OANN achieves multiclass classification accuracies exceeding 90 %, comparable to conventional machine learning approaches. This proof-of-concept demonstrates the feasibility of a multilayer OANN in the frequency domain and can be extended to a scalable, integrated photonic platform with ultrafast weights updates, with potential applications to single-shot classification in spectroscopy.

AB - Optical artificial neural networks (OANNs) leverage the advantages of photonic technologies including high processing speeds, low energy consumption, and mass production to establish a competitive and scalable platform for machine learning applications. While recent advancements have focused on harnessing spatial or temporal modes of light, the frequency domain attracts a lot of attention, with current implementations including spectral multiplexing, neural networks in nonlinear optical systems and extreme learning machines. Here, we present an experimental realization of a programmable photonic frequency circuit, realized with fiber-optical components, and implement the in-situ training with optical weight control of an OANN operating in the frequency domain. Input data is encoded into phases of frequency comb modes, and programmable phase and amplitude manipulations of the spectral modes enable in-situ training of the OANN, without employing a digital model of the device. The trained OANN achieves multiclass classification accuracies exceeding 90 %, comparable to conventional machine learning approaches. This proof-of-concept demonstrates the feasibility of a multilayer OANN in the frequency domain and can be extended to a scalable, integrated photonic platform with ultrafast weights updates, with potential applications to single-shot classification in spectroscopy.

KW - machine learning

KW - photonic computing

KW - ultrafast optics

UR - http://www.scopus.com/inward/record.url?scp=105009132590&partnerID=8YFLogxK

U2 - 10.1515/nanoph-2025-0125

DO - 10.1515/nanoph-2025-0125

M3 - Article

AN - SCOPUS:105009132590

VL - 14

SP - 2779

EP - 2786

JO - Nanophotonics

JF - Nanophotonics

SN - 2192-8606

IS - 16

ER -

Von denselben Autoren