EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Synthetic Spectrogram Augmentation via Semi-Supervised WGAN-GP for Acoustic Industrial Quality Inspection of Turbine Housings Under Extreme Data Scarcity

Ander Gracia Moisés iD, Óscar Del Barrio Farran iD, David Martinez García iD, María Puy Zudaire Latienda iD

Sensors · 12.05.2026 · 26 (10) · 3052

DOI: 10.3390/s26103052 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 12.05.2026. Verfasst von Ander Gracia Moisés und 3 weiteren Autor:innen. Beteiligte Einrichtung: NAITEC-Fundación I + D Automoción y Mecatrónica, Calle Tajonar 20, 31006 Pamplona, Spain. Die Arbeit nennt 27 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Gobierno de Navarra.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Impact-based acoustic inspection provides a rapid non-destructive approach for screening metallic components by analyzing the sound radiated after a controlled mechanical excitation. However, the limited availability of labeled data from defective parts remains a major challenge for deploying deep learning classifiers in production. This paper proposes a complete pipeline that converts raw impact-response audio recordings into magnitude log-spectrogram images and trains a semi-supervised Wasserstein GAN with gradient penalty (SS-WGAN-GP) designed to operate under extreme data scarcity. The architecture couples a shared convolutional backbone with two output heads: a Wasserstein critic for unsupervised discrimination between real and generated samples, and a binary classification head for supervised quality labeling, jointly optimized through a combined loss that balances Wasserstein distance, gradient penalty, and cross-entropy. A key property of the design is that the generator acts as a source of synthetic training samples, producing progressively more realistic spectrograms as training advances. These samples, in turn, enrich the feature representations learned by the shared backbone and improve the performance of the classification head. The classification head of the trained discriminator is deployed directly as the quality classifier, without requiring external data or post hoc retraining.

Schlagwörter

Stichwörter der Autor:innen

acoustic NDTaudio spectrogramsdata augmentationgenerative adversarial networksindustrial quality inspectionsemi-supervised learningturbine housings

Autor:innen und Institutionen

  • NAITEC-Fundación I + D Automoción y Mecatrónica, Calle Tajonar 20, 31006 Pamplona, Spain.

ORCID-Profile: Ander Gracia Moisés, Óscar Del Barrio Farran, David Martinez García, María Puy Zudaire Latienda

Förderung

  • Gobierno de Navarra

Zitieren

APA

Moisés, A. G., Farran, Ó. D. B., García, D. M., & Latienda, M. P. Z. (2026). Synthetic Spectrogram Augmentation via Semi-Supervised WGAN-GP for Acoustic Industrial Quality Inspection of Turbine Housings Under Extreme Data Scarcity. Sensors, 26(10), 3052. https://doi.org/10.3390/s26103052

Vancouver

Moisés AG, Farran ÓDB, García DM, Latienda MPZ. Synthetic Spectrogram Augmentation via Semi-Supervised WGAN-GP for Acoustic Industrial Quality Inspection of Turbine Housings Under Extreme Data Scarcity. Sensors. 2026;26(10):3052. doi:10.3390/s26103052

BibTeX
@article{Moiss2026,
  author = {Moisés, Ander Gracia and Farran, Óscar Del Barrio and García, David Martinez and Latienda, María Puy Zudaire},
  title = {Synthetic Spectrogram Augmentation via Semi-Supervised WGAN-GP for Acoustic Industrial Quality Inspection of Turbine Housings Under Extreme Data Scarcity},
  journal = {Sensors},
  year = {2026},
  volume = {26},
  number = {10},
  pages = {3052},
  doi = {10.3390/s26103052}
}
Kostenlos starten →