Spectral library transfer between distinct laser-induced
breakdown spectroscopy systems trained on simultaneous
measurements

J 2023

Spectral library transfer between distinct laser-induced breakdown spectroscopy systems trained on simultaneous measurements

VRÁBEL, Jakub, Erik KÉPEŠ, Pavel NEDĚLNÍK, Jakub BUDAY, Jan CEMPÍREK et. al.

Základní údaje

Originální název

Spectral library transfer between distinct laser-induced breakdown spectroscopy systems trained on simultaneous measurements

Autoři

VRÁBEL, Jakub, Erik KÉPEŠ, Pavel NEDĚLNÍK, Jakub BUDAY, Jan CEMPÍREK (203 Česká republika, domácí), Pavel POŘÍZKA a Jozef KAISER

Vydání

Journal of Analytical Atomic Spectrometry, Royal Society of Chemistry, 2023, 0267-9477

Další údaje

Jazyk

angličtina

Typ výsledku

Článek v odborném periodiku

Obor

10505 Geology

Stát vydavatele

Velká Británie a Severní Irsko

Utajení

není předmětem státního či obchodního tajemství

Odkazy

URL

Impakt faktor

Impact factor: 3.400 v roce 2022

Kód RIV

RIV/00216224:14310/23:00132292

Organizační jednotka

Přírodovědecká fakulta

DOI

http://dx.doi.org/10.1039/d2ja00406b

UT WoS

000940533400001

Klíčová slova anglicky

spectroscopic data; library transfer; machine learning; artificial neural networks; autoencoder

Štítky

rivok

Příznaky

Mezinárodní význam, Recenzováno

Změněno: 16. 11. 2023 10:02, Mgr. Marie Šípková, DiS.

Anotace

V originále

The mutual incompatibility of distinct spectroscopic systems is among the most limiting factors in laser-induced breakdown spectroscopy (LIBS). The cost related to setting up a new LIBS system is increased, as its extensive calibration is required. Solving this problem would enable inter-laboratory reference measurements and shared spectral libraries, which are fundamental for other spectroscopic techniques. We study a simplified version of this challenge where LIBS systems differ only in the spectrometers used and collection optics but share all other parts of the apparatus and collect spectra simultaneously from the same plasma plume. Extensive datasets measured as hyperspectral images of a heterogeneous rock sample are used to train machine learning models that can transfer spectra between systems. The transfer is realized using a composed model that consists of a variational autoencoder (VAE) and a multilayer perceptron (MLP). The VAE is used to create a latent representation of spectra from a primary system. Subsequently, spectra from a secondary system are mapped to corresponding locations in the latent space by the MLP. The transfer is evaluated using several figures of merit (Euclidean and cosine distances, both spatially resolved; k-means clustering of transferred spectra). We demonstrate the viability of the method and compare it to several baseline approaches of varying complexities.

Citovat

VRÁBEL, Jakub, Erik KÉPEŠ, Pavel NEDĚLNÍK, Jakub BUDAY, Jan CEMPÍREK, Pavel POŘÍZKA a Jozef KAISER. Spectral library transfer between distinct laser-induced breakdown spectroscopy systems trained on simultaneous measurements. Journal of Analytical Atomic Spectrometry. Royal Society of Chemistry, 2023, roč. 38, č. 4, s. 841-853. ISSN 0267-9477. Dostupné z: https://dx.doi.org/10.1039/d2ja00406b.

@article{2339937,
   author = {Vrábel, Jakub and Képeš, Erik and Nedělník, Pavel and Buday, Jakub and Cempírek, Jan and Pořízka, Pavel and Kaiser, Jozef},
   article_number = {4},
   doi = {http://dx.doi.org/10.1039/d2ja00406b},
   keywords = {spectroscopic data; library transfer; machine learning; artificial neural networks; autoencoder},
   language = {eng},
   issn = {0267-9477},
   journal = {Journal of Analytical Atomic Spectrometry},
   title = {Spectral library transfer between distinct laser-induced breakdown spectroscopy systems trained on simultaneous measurements},
   url = {https://doi.org/10.1039/D2JA00406B},
   volume = {38},
   year = {2023}
}

TY  - JOUR
ID  - 2339937
AU  - Vrábel, Jakub - Képeš, Erik - Nedělník, Pavel - Buday, Jakub - Cempírek, Jan - Pořízka, Pavel - Kaiser, Jozef
PY  - 2023
TI  - Spectral library transfer between distinct laser-induced breakdown spectroscopy systems trained on simultaneous measurements
JF  - Journal of Analytical Atomic Spectrometry
VL  - 38
IS  - 4
SP  - 841-853
EP  - 841-853
PB  - Royal Society of Chemistry
SN  - 02679477
KW  - spectroscopic data
KW  - library transfer
KW  - machine learning
KW  - artificial neural networks
KW  - autoencoder
UR  - https://doi.org/10.1039/D2JA00406B
N2  - The mutual incompatibility of distinct spectroscopic systems is among the most limiting factors in laser-induced breakdown spectroscopy (LIBS). The cost related to setting up a new LIBS system is increased, as its extensive calibration is required. Solving this problem would enable inter-laboratory reference measurements and shared spectral libraries, which are fundamental for other spectroscopic techniques. We study a simplified version of this challenge where LIBS systems differ only in the spectrometers used and collection optics but share all other parts of the apparatus and collect spectra simultaneously from the same plasma plume. Extensive datasets measured as hyperspectral images of a heterogeneous rock sample are used to train machine learning models that can transfer spectra between systems. The transfer is realized using a composed model that consists of a variational autoencoder (VAE) and a multilayer perceptron (MLP). The VAE is used to create a latent representation of spectra from a primary system. Subsequently, spectra from a secondary system are mapped to corresponding locations in the latent space by the MLP. The transfer is evaluated using several figures of merit (Euclidean and cosine distances, both spatially resolved; k-means clustering of transferred spectra). We demonstrate the viability of the method and compare it to several baseline approaches of varying complexities.
ER  -

VRÁBEL, Jakub, Erik KÉPEŠ, Pavel NEDĚLNÍK, Jakub BUDAY, Jan CEMPÍREK, Pavel POŘÍZKA a Jozef KAISER. Spectral library transfer between distinct laser-induced breakdown spectroscopy systems trained on simultaneous measurements. \textit{Journal of Analytical Atomic Spectrometry}. Royal Society of Chemistry, 2023, roč.~38, č.~4, s.~841-853. ISSN~0267-9477. Dostupné z: https://dx.doi.org/10.1039/d2ja00406b.

Podrobný výpis o publikaci