EUVIMEDEUVIMED entdecken
Crossref

Tinybionet: A lightweight time-frequency network for biomedical signal classification on edge devices

Sandra Stankovic iD, Stevica Cvetkovic iD, Sasa V Nikolic iD

Technology and Health Care · 04.08.2026

DOI: 10.1177/09287329261473368 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Technology and Health Care, online erschienen am 04.08.2026. Verfasst von Sandra Stankovic und 2 weiteren Autor:innen. Beteiligte Einrichtung: Faculty of Electronic Engineering, University of Nis, Nis, Serbia. Die Arbeit nennt 53 Literaturangaben. Gefördert von: Ministarstvo Prosvete, Nauke i Tehnološkog Razvoja.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Real-time analysis of biomedical time-series on edge devices faces significant challenges due to signal non-stationarity, environmental noise, and the strict computational limitations of embedded hardware. While deep learning offers superior accuracy, standard architectures are often too computationally demanding for embedded deployment. In this work, we propose TinyBioNet, a highly compact convolutional neural network designed explicitly for the efficient affective state classification on devices with limited resources. Unlike traditional 1D CNNs that implicitly learn spectral features, or lightweight architectures that treat the STFT as an isolated preprocessing operation, our approach natively embeds a fixed-basis time-frequency transformation directly within the network execution graph using parallel 1D convolutional layers. This design eliminates separate preprocessing stages, reduces runtime, and allows a highly compact backend 2D CNN topology to operate immediately on structured complex time-frequency feature maps. By combining this structured input with depthwise convolutions and residual connections, TinyBioNet achieves state-of-the-art performance with only 5.6k parameters. We further optimize the model for embedded targets through aggressive low-bit quantization. Comprehensive evaluations on three public datasets demonstrate robust generalization across various biomedical signal modalities. The proposed framework achieves classification accuracies of up to 98.89% and 99.38% for PPG and ACC signals, respectively, while maintaining negligible performance degradation under 4-bit integer quantization.

Schlagwörter

Stichwörter der Autor:innen

affective computingbiomedical signalslightweight CNNquantizationtime-frequency analysis

Autor:innen und Institutionen

  • Faculty of Electronic Engineering, University of Nis, Nis, Serbia.

ORCID-Profile: Sandra Stankovic, Stevica Cvetkovic, Sasa V Nikolic

Förderung

  • Ministarstvo Prosvete, Nauke i Tehnološkog Razvoja

Zitieren

APA

Stankovic, S., Cvetkovic, S., & Nikolic, S. V. (2026). Tinybionet: A lightweight time-frequency network for biomedical signal classification on edge devices. Technology and Health Care. https://doi.org/10.1177/09287329261473368

Vancouver

Stankovic S, Cvetkovic S, Nikolic SV. Tinybionet: A lightweight time-frequency network for biomedical signal classification on edge devices. Technology and Health Care. 2026. doi:10.1177/09287329261473368

BibTeX
@article{Stankovic2026,
  author = {Stankovic, Sandra and Cvetkovic, Stevica and Nikolic, Sasa V},
  title = {Tinybionet: A lightweight time-frequency network for biomedical signal classification on edge devices},
  journal = {Technology and Health Care},
  year = {2026},
  doi = {10.1177/09287329261473368}
}
Kostenlos starten →