講座
Journal of the Spectroscopical Society of Japan · 2024 · 73 (4) · 113
DOI: 10.5111/bunkou.73.4_113 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of the Spectroscopical Society of Japan, 2024. Verfasst von 誠史 福原.
Abstract (Originalsprache)
Spectrum analysis with convolutional neural networks (CNNs) has attracted much attention since breakthrough of deeplearning technology in image recognition field.However, CNNs are still considered as a "black box" while they provide impressive and unique applications in many fields, including optical spectrum analysis.In this article, examples how CNN focus on spectrum features depending on parameters are explained by analyzing simple Lorentzian spectra and Raman spectra obtained from pharmaceutical compounds and amino acids.
Zitieren
APA
福原, 誠. (2024). 講座. Journal of the Spectroscopical Society of Japan, 73(4), 113. https://doi.org/10.5111/bunkou.73.4_113
Vancouver
福原 誠. 講座. Journal of the Spectroscopical Society of Japan. 2024;73(4):113. doi:10.5111/bunkou.73.4_113
BibTeX
@article{2024,
author = {福原, 誠史},
title = {講座},
journal = {Journal of the Spectroscopical Society of Japan},
year = {2024},
volume = {73},
number = {4},
pages = {113},
doi = {10.5111/bunkou.73.4_113}
}
EUVIMED