The Impact of Frequency Bands on Acoustic Anomaly Detection of Machines using Deep Learning Based Model (2403.00379v1)
Abstract: In this paper, we propose a deep learning based model for Acoustic Anomaly Detection of Machines, the task for detecting abnormal machines by analysing the machine sound. By conducting extensive experiments, we indicate that multiple techniques of pseudo audios, audio segment, data augmentation, Mahalanobis distance, and narrow frequency bands, which mainly focus on feature engineering, are effective to enhance the system performance. Among the evaluating techniques, the narrow frequency bands presents a significant impact. Indeed, our proposed model, which focuses on the narrow frequency bands, outperforms the DCASE baseline on the benchmark dataset of DCASE 2022 Task 2 Development set. The important role of the narrow frequency bands indicated in this paper inspires the research community on the task of Acoustic Anomaly Detection of Machines to further investigate and propose novel network architectures focusing on the frequency bands.
- “Vae-based anomaly detection with domain adaptation,” Tech. Rep., DCASE Challenge, 2021.
- “Dadaed-double anomaly detector with aediff,” Tech. Rep., DCASE Challenge, 2022.
- Shuxian Wang et al, “Ensemble of multiple anomaly detectors under domain generalization conditions,” Tech. Rep., DCASE Challenge, 2022.
- “Two-stage anomalous sound detection systems using domain generalization and specialization techniques,” Tech. Rep., DCASE Challenge, 2022.
- Kevin Wilkinghoff, “An outlier exposed anomalous sound detection system for domain generalization in machine condition monitoring,” Tech. Rep., DCASE Challenge, 2022.
- “Unsupervised anomalous sound detection using multiple time-frequency representations,” Tech. Rep., DCASE Challenge, 2022.
- “Dcase challenge 2022: Self-supervised learning pre-training, training for unsupervised anomalous sound detection,” Tech. Rep., DCASE Challenge, 2022.
- “Robust anomaly sound detection framework for machine condition monitoring,” Tech. Rep., DCASE Challenge, 2022.
- “Comparative experiments on spectrogram representation for anomalous sound detection,” Tech. Rep., DCASE Challenge, 2022.
- “Outlier detection algorithm based on gaussian mixture model,” in Proc. ICPICS, 2019, pp. 488–492.
- “Lof: identifying density-based local outliers,” in Proc. ACM SIGMOD, 2000, pp. 93–104.
- “Mobilenetv2: Inverted residuals and linear bottlenecks,” in Proc. CVPR, 2018, pp. 4510–4520.
- “Euclidean distance matrices: essential theory, algorithms, and applications,” IEEE Signal Processing Magazine, vol. 32, no. 6, pp. 12–30, 2015.
- “A novel multi-window spectrogram augmentation approach for speech emotion recognition using deep learning,” in Proc. ICMEAS, 2021, pp. 1–6.
- “Emotion speech synthesis method based on multi-channel time–frequency domain generative adversarial networks (mc-tfd gans) and mixup,” Arabian Journal for Science and Engineering, vol. 47, no. 2, pp. 1749–1762, 2022.
- “Blade fault diagnosis using mahalanobis distance,” Journal of Mechanical Science and Technology, vol. 35, pp. 1377–1385, 2021.
- Kota Dohi et al, “Description and discussion on DCASE 2022 challenge task 2: Unsupervised anomalous sound detection for machine condition monitoring applying domain generalization techniques,” in Proc. DCASE, 2022, pp. 1–5.
- Kota Dohi et al, “MIMII DG: Sound dataset for malfunctioning industrial machine investigation and inspection for domain generalization task,” in Proc. DCASE, 2022, pp. 1–5.
- Noboru Harada et al, “ToyADMOS2: Another dataset of miniature-machine operating sounds for anomalous sound detection under domain shift conditions,” in Proc. DCASE, 2021, pp. 1–5.
- P. K. Diederik and B. Jimmy, “Adam: A method for stochastic optimization,” CoRR, vol. abs/1412.6980, 2015.