Physics-Aware Generative Demasking: Spatially Conditioned Diffusion for Robust Transient Detection in Industrial Noise
Entropy · 24.03.2026 · 28 (4) · 364
DOI: 10.3390/e28040364 ↗Kurz gefasst
Fachartikel in der Zeitschrift Entropy, online erschienen am 24.03.2026. Verfasst von Hailin Cao und 5 weiteren Autor:innen. Beteiligte Einrichtung: School of Microelectronics and Communication Engineering, Chongqing University, Chongqing 400044, China. Die Arbeit nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Chongqing Natural Science Foundation Innovation and Development Joint Fund.
Abstract (Originalsprache)
Detecting transient “click” sounds during connector insertion is pivotal for automotive assembly quality but remains intractable due to high-intensity, non-stationary industrial noise. This paper introduces a physics-aware generative demasking framework that integrates acoustic spatial priors with conditional diffusion modeling. We propose the spatially conditioned diffusion probabilistic model (SC-DPM), where an ambient reference signal acts as a physical constraint to steer the reverse diffusion process. By exploiting the spatial decay of insertion sounds, this mechanism effectively disentangles the target transient from the background noise manifold, reconstructing high-fidelity spectro-temporal features. Discriminative temporal patterns are extracted using causal random convolutional kernels with causal dilations and local proportion of positive values (LPPV) pooling. Experiments on real-world datasets demonstrate 93.3% accuracy. The proposed “restore-then-classify” paradigm significantly enhances robustness against acoustic variability, establishing a scalable methodology for precise industrial monitoring under extreme noise conditions.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- School of Microelectronics and Communication Engineering, Chongqing University, Chongqing 400044, China.
- Great Wall Motor Company Limited, Chongqing Branch, No. 666, Fenglong Avenue, Yongchuan District, Chongqing 402100, China.
ORCID-Profile: Hailin Cao
Förderung
- Chongqing Natural Science Foundation Innovation and Development Joint Fund
Zitieren
APA
Cao, H., Lv, Z., Hu, J., Wang, H., Yang, L., & Zhang, G. (2026). Physics-Aware Generative Demasking: Spatially Conditioned Diffusion for Robust Transient Detection in Industrial Noise. Entropy, 28(4), 364. https://doi.org/10.3390/e28040364
Vancouver
Cao H, Lv Z, Hu J, Wang H, Yang L, Zhang G. Physics-Aware Generative Demasking: Spatially Conditioned Diffusion for Robust Transient Detection in Industrial Noise. Entropy. 2026;28(4):364. doi:10.3390/e28040364
BibTeX
@article{Cao2026,
author = {Cao, Hailin and Lv, Zixi and Hu, Jinjie and Wang, Hui and Yang, Lisheng and Zhang, Guoxin},
title = {Physics-Aware Generative Demasking: Spatially Conditioned Diffusion for Robust Transient Detection in Industrial Noise},
journal = {Entropy},
year = {2026},
volume = {28},
number = {4},
pages = {364},
doi = {10.3390/e28040364}
}
EUVIMED