EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

HyperKAN: Kolmogorov–Arnold Networks Make Hyperspectral Image Classifiers Smarter

Nikita Firsov iD, Evgeny Myasnikov, Valeriy Lobanov iD, Roman Khabibullin iD, Nikolay Kazanskiy iD, Svetlana Khonina iD, Muhammad A Butt iD, Artem Nikonorov

Sensors · 30.11.2024 · 24 (23) · 7683

DOI: 10.3390/s24237683 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 30.11.2024. Verfasst von Nikita Firsov und 7 weiteren Autor:innen. Beteiligte Einrichtung: Samara National Research University, Samara 443086, Russia. Die Arbeit wurde bisher 56-mal zitiert (Crossref) und nennt 43 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Analytical Center for the Government of the Russian Federation.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

In traditional neural network designs, a multilayer perceptron (MLP) is typically employed as a classification block following the feature extraction stage. However, the Kolmogorov–Arnold Network (KAN) presents a promising alternative to MLP, offering the potential to enhance prediction accuracy. In this paper, we studied KAN-based networks for pixel-wise classification of hyperspectral images. Initially, we compared baseline MLP and KAN networks with varying numbers of neurons in their hidden layers. Subsequently, we replaced the linear, convolutional, and attention layers of traditional neural networks with their KAN-based counterparts. Specifically, six cutting-edge neural networks were modified, including 1D (1DCNN), 2D (2DCNN), and 3D convolutional networks (two different 3DCNNs, NM3DCNN), as well as transformer (SSFTT). Experiments conducted using seven publicly available hyperspectral datasets demonstrated a substantial improvement in classification accuracy across all the networks. The best classification quality was achieved using a KAN-based transformer architecture.

Schlagwörter

Stichwörter der Autor:innen

Kolmogorov–Arnold networksclassificationconvolutional neural networkshyperspectral imagingtransformers

Autor:innen und Institutionen

  • Samara National Research University, Samara 443086, Russia.
  • Adyghe State University, Maykop 385000, Russia.

ORCID-Profile: Nikita Firsov, Valeriy Lobanov, Roman Khabibullin, Nikolay Kazanskiy, Svetlana Khonina, Muhammad A Butt

Förderung

  • Analytical Center for the Government of the Russian Federation

Zitieren

APA

Firsov, N., Myasnikov, E., Lobanov, V., Khabibullin, R., Kazanskiy, N., Khonina, S., Butt, M. A., & Nikonorov, A. (2024). HyperKAN: Kolmogorov–Arnold Networks Make Hyperspectral Image Classifiers Smarter. Sensors, 24(23), 7683. https://doi.org/10.3390/s24237683

Vancouver

Firsov N, Myasnikov E, Lobanov V, Khabibullin R, Kazanskiy N, Khonina S, et al. HyperKAN: Kolmogorov–Arnold Networks Make Hyperspectral Image Classifiers Smarter. Sensors. 2024;24(23):7683. doi:10.3390/s24237683

BibTeX
@article{Firsov2024,
  author = {Firsov, Nikita and Myasnikov, Evgeny and Lobanov, Valeriy and Khabibullin, Roman and Kazanskiy, Nikolay and Khonina, Svetlana and Butt, Muhammad A and Nikonorov, Artem},
  title = {HyperKAN: Kolmogorov–Arnold Networks Make Hyperspectral Image Classifiers Smarter},
  journal = {Sensors},
  year = {2024},
  volume = {24},
  number = {23},
  pages = {7683},
  doi = {10.3390/s24237683}
}
Kostenlos starten →