EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Design and Evaluation of a Compact VGG-Inspired CNN for Keyword Spotting in Resource-Constrained TinyML Systems

Wilson Gustavo Chango, Mayra Barrera, Daniel Maldonado-Ruiz, Julio Balarezo, Marcelo V. Garcia, Geovanny Silva

Computation · 13.05.2026 · 14 (5) · 112

DOI: 10.3390/computation14050112 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Computation, online erschienen am 13.05.2026. Verfasst von Wilson Gustavo Chango und 5 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 46 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This paper investigates the design and evaluation of compact convolutional neural networks (CNNs) for keyword spotting (KWS) and acoustic event detection under the stringent constraints of the TinyML paradigm. The research expands upon traditional binary classification approaches by addressing a multi-class acoustic scenario encompassing eight distinct categories: stop, no, go, yes, unknown, silence, noise_ambient, and noise_sudden. The primary objective is to evaluate the feasibility of deploying reliable acoustic detection systems on ultra-low-power microcontrollers for edge computing applications. To this end, five lightweight architectures were developed and benchmarked: AlexNet-Tiny, LeNet-Tiny, MobileNet-Tiny, VGG-Tiny, and CustomCNN-Tiny. The models were trained using Mel-spectrogram features and optimized through INT8 post-training quantization to facilitate embedded deployment. Hardware simulation was conducted targeting the XIAO nRF52840 Sense microcontroller (64 MHz, 256 KB RAM). Experimental results demonstrate that the Gold VGG-Tiny architecture achieves the highest classification accuracy (89.81%), while Silver MobileNet-Tiny provides the superior operational efficiency with the lowest inference latency (0.88 ms) and minimal energy consumption (14.4 µJ). Furthermore, the Bronze CustomCNN-Tiny model achieves the most reduced memory footprint (42.9 KB), highlighting its suitability for memory-constrained environments. Statistical validation using Cohen’s Kappa, Matthews Correlation Coefficient (MCC), and Area Under the Curve (AUC) confirms the robustness and reliability of the proposed models. The potential application of this system is motivated by acoustic monitoring for the early detection of high-risk situations, such as gender-based violence. Future work will focus on on-device physical validation and real-world deployment in wearable safety electronics.

Zitieren

APA

Chango, W. G., Barrera, M., Maldonado-Ruiz, D., Balarezo, J., Garcia, M. V., & Silva, G. (2026). Design and Evaluation of a Compact VGG-Inspired CNN for Keyword Spotting in Resource-Constrained TinyML Systems. Computation, 14(5), 112. https://doi.org/10.3390/computation14050112

Vancouver

Chango WG, Barrera M, Maldonado-Ruiz D, Balarezo J, Garcia MV, Silva G. Design and Evaluation of a Compact VGG-Inspired CNN for Keyword Spotting in Resource-Constrained TinyML Systems. Computation. 2026;14(5):112. doi:10.3390/computation14050112

BibTeX
@article{Chango2026,
  author = {Chango, Wilson Gustavo and Barrera, Mayra and Maldonado-Ruiz, Daniel and Balarezo, Julio and Garcia, Marcelo V. and Silva, Geovanny},
  title = {Design and Evaluation of a Compact VGG-Inspired CNN for Keyword Spotting in Resource-Constrained TinyML Systems},
  journal = {Computation},
  year = {2026},
  volume = {14},
  number = {5},
  pages = {112},
  doi = {10.3390/computation14050112}
}
Kostenlos starten →