Multi-scale Scanning Network for Machine Anomalous Sound Detection

2025-08-23

AI Summary

To address insufficient modeling of multi-scale time-frequency patterns in machine anomaly sound detection, this paper proposes the Multi-Scale Scanning Network (MS-Net). MS-Net employs parallel convolutional kernels of varying sizes to scan Mel-spectrogram inputs, explicitly capturing recurrent time-frequency structures across scales. A lightweight, shared-weight convolutional design is introduced to reduce parameter count while enhancing generalization and scalability. This work constitutes the first systematic end-to-end modeling of regular multi-scale acoustic patterns inherent in industrial machinery sounds for anomaly detection. Evaluated on the DCASE 2020 and 2023 Task 2 benchmarks, MS-Net achieves state-of-the-art performance, demonstrating significant robustness to heterogeneous machine types and noisy acoustic environments. The results empirically validate the critical importance of multi-scale structural modeling for industrial acoustic anomaly detection.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Data Mining & Knowledge Management: Anomaly/Outlier Detection

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• Web Mining and Content Analysis: Machine learning and data science for the Web
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Machine sounds exhibit consistent and repetitive patterns in both the frequency and time domains, which vary significantly across scales for different machine types. For instance, rotating machines often show periodic features in short time intervals, while reciprocating machines exhibit broader patterns spanning the time domain. While prior studies have leveraged these patterns to improve Anomalous Sound Detection (ASD), the variation of patterns across scales remains insufficiently explored. To address this gap, we introduce a Multi-scale Scanning Network (MSN) designed to capture patterns at multiple scales. MSN employs kernel boxes of varying sizes to scan audio spectrograms and integrates a lightweight convolutional network with shared weights for efficient and scalable feature representation. Experimental evaluations on the DCASE 2020 and DCASE 2023 Task 2 datasets demonstrate that MSN achieves state-of-the-art performance, highlighting its effectiveness in advancing ASD systems.

Problem

• Detecting anomalous machine sounds across varying frequency and time scales
• Capturing multi-scale patterns in audio spectrograms for different machine types
• Improving anomalous sound detection performance through scalable feature representation

Innovation

• Multi-scale Scanning Network captures patterns
• Varying kernel boxes scan audio spectrograms
• Lightweight convolutional network with shared weights