EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Progressive Training of Hybrid Clipped Rectified Linear Unit for Resilient Convolution Neural Networks

Seyedhamidreza Mousavi, Mohammad Hasan Ahmadilivani, Jaan Raik, Maksim Jenihhin, Masoud Daneshtalab

IEEE Access · 2026 · 14 · 39033-39045

DOI: 10.1109/access.2026.3671327 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Access, 2026. Verfasst von Seyedhamidreza Mousavi und 4 weiteren Autor:innen. Die Arbeit nennt 47 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Academic Infrastructure for Supercomputing in Sweden (NAISS) funded by the Swedish Research Council, European Union (EU) through “TAICHIP”, EU through NexTArc.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Kein Abstract verfügbar.

Förderung

  • National Academic Infrastructure for Supercomputing in Sweden (NAISS) funded by the Swedish Research Council
  • European Union (EU) through “TAICHIP”
  • EU through NexTArc
  • Estonian Research Council “Cross-Layer Reliability and Self-Health Awareness for Intelligent Autonomous Systems (CRASHLESS)”
  • CoE TK202 “Universum.”

Zitieren

APA

Mousavi, S., Ahmadilivani, M. H., Raik, J., Jenihhin, M., & Daneshtalab, M. (2026). Progressive Training of Hybrid Clipped Rectified Linear Unit for Resilient Convolution Neural Networks. IEEE Access, 14, 39033-39045. https://doi.org/10.1109/access.2026.3671327

Vancouver

Mousavi S, Ahmadilivani MH, Raik J, Jenihhin M, Daneshtalab M. Progressive Training of Hybrid Clipped Rectified Linear Unit for Resilient Convolution Neural Networks. IEEE Access. 2026;14:39033-39045. doi:10.1109/access.2026.3671327

BibTeX
@article{Mousavi2026,
  author = {Mousavi, Seyedhamidreza and Ahmadilivani, Mohammad Hasan and Raik, Jaan and Jenihhin, Maksim and Daneshtalab, Masoud},
  title = {Progressive Training of Hybrid Clipped Rectified Linear Unit for Resilient Convolution Neural Networks},
  journal = {IEEE Access},
  year = {2026},
  volume = {14},
  pages = {39033--39045},
  doi = {10.1109/access.2026.3671327}
}
Kostenlos starten →