28 August 2023 in IEEE Transactions on Systems, Man, and Cybernetics: Systems
Large monitoring systems produce data that is often compressed to be transmitted over the network. For latency or security reasons, compressed data may be processed at the edge, i.e., along the path from sensors to the cloud, for some purposes such as anomaly detection. However, the performance of a detector distinguishing between normal and anomalous behavior may be affected by the loss of information due to compression. We here analyze how lossy compression affects the performance of a generic anomaly detector. This relationship is formalized in terms of information-theoretic quantities. Within such a framework we leverage a Gaussian assumption to derive analytical results regarding the importance of white noise as a representative of both the average and asymptotic anomalies. Moreover, in an anomaly-agnostic scenario, we also show the existence of a level of compression for which an anomaly is undetectable though compression is not completely destructive. Numerical evidence confirms that the proposed information-theoretic quantities anticipate the performance of practical compressors and detectors in the case of Gaussian and non-Gaussian signals allowing an assessment of the tradeoff between compression and detection.
3 July 2024 in IEEE Transactions on Industrial Informatics
Anomalies are rare events, and this affects the design flow of detectors that monitor systems that behave normally most of the time but whose failure may have serious consequences. This limitation is particularly evident in the detector performance evaluation: it requires an abundance of normal and anomalous data but realistically faces a scarcity of the latter. To address this, in this article, we develop a framework comprising a set of abstract anomalies modeling the effects real-world failures and disturbances have on sensor readings. In addition, we devise synthetic generation procedures for these anomalies. Given a dataset of normal tracks from the actual application, one may apply such procedures to produce anomalous-like time series for a comprehensive detector assessment. We show that this framework can anticipate the detector performing best with real-world anomalies in the context of human and structural health monitoring, also highlighting that, in these cases, the best detector is not the most complex.
2 August 2026 in SSRN Electronic Journal
In this work, we introduce and validate a Compressed Sensing (CS) framework specifically tailored for vibrationbased Structural Health Monitoring (SHM), aimed at reducing local processing, transmission, and storage costs while preserving the information required for Operational Modal Analysis (OMA). On the encoding side, we extend Modelassisted Rakeness-based CS (MRak) by proposing Adapted MRak (AMRak), which designs the sensing mechanismthrough an amplitude-aware modal-band mask driven by prior modal information to protect weak yet diagnostically relevant spectral components under high compression. On the decoding side, we propose a support-aware neural reconstruction pipeline termed Support-aware Refinement and Debiasing (SRD): reference supports are constructed offline within the modal bands through an energy-based selection rule and are used to guide the training loss of a transform-domain multi-channel U-Net. At inference time, compressed measurements are first inverted through a coarse measurement-consistent step and then refined by the network; the refined coefficients are used to re-estimate the dominant support, enabling a final support-restricted least-squares debiasing stage that corrects peak-amplitudeattenuation and suppresses out-of-band content. The reconstructed signals enable reliable OMA using different approaches, including non-parametric Frequency Domain Decomposition and parametric Stochastic Subspace Identification with automated pole selection. Experiments on the publicly available Bridge S101 benchmark and on an additional anonymized multi-sensor monitoring dataset show that the proposed AMRak+SRD framework preserves modal peak fidelity and mode shape consistency at compression ratios up to 10, a regime in which generic CS methods typically fail.
12 October 2021 in TechRxiv
We analyze the effect of lossy compression in the processing of sensor signals that must be used to detect anomalous events in the system under observation. The intuitive relationship between the quality loss at higher compression and the possibility of telling anomalous behaviours from normal ones is formalized in terms of information-theoretic quantities. Some analytic derivations are made within the Gaussian framework and possibly in the asymptotic regime for what concerns the stretch of signals considered. Analytical conclusions are matched with the performance of practical detectors in a toy case allowing the assessment of different compression/detector configurations.
15 May 2025 in IEEE Transactions on Pattern Analysis and Machine Intelligence
As neural networks are trained to perform tasks of increasing complexity, their size increases, which presents several challenges in their deployment on devices with limited resources. To cope with this, a recently proposed approach hinges on substituting the classical Multiply-and-ACcumulate (MAC) neurons in the hidden layers with other neurons called Multiply-And-Max/min (MAM) whose selective behavior helps identify important interconnections, thus allowing aggressive pruning of the others. Hybrid MAM&MAC structures promise a 10x or even 100x reduction in their memory footprint compared to what can be obtained by pruning MAC-only structures. However, a cornerstone of maintaining this promise is the assumption that MAC&MAM architectures have the same expressive power as MAC-only ones. To concretize such a cornerstone, we take here a step in the theoretical characterization of the capabilities of mixed MAM&MAC networks. We prove, with two theorems, that two hidden MAM layers followed by a MAC neuron with possibly a normalization stage is a universal approximator.