ASDNet: An Efficient Self-Supervised Convolutional Network for Anomalous Sound Detection
Anomalous Sound Detection (ASD) is crucial for ensuring industrial equipment safety and enhancing production efficiency. However, existing methods, while pursuing high detection accuracy, are often associated with high computational complexity, making them unsuitable for resource-constrained environ...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-01-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/15/2/584 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832589287373144064 |
|---|---|
| author | Dewei Kong Guoshun Yuan Hongjiang Yu Shuai Wang Bo Zhang |
| author_facet | Dewei Kong Guoshun Yuan Hongjiang Yu Shuai Wang Bo Zhang |
| author_sort | Dewei Kong |
| collection | DOAJ |
| description | Anomalous Sound Detection (ASD) is crucial for ensuring industrial equipment safety and enhancing production efficiency. However, existing methods, while pursuing high detection accuracy, are often associated with high computational complexity, making them unsuitable for resource-constrained environments. This study proposes an efficient self-supervised ASD framework that integrates spectral features, lightweight neural networks, and various anomaly scoring methods. Unlike traditional Log-Mel features, spectral features retain richer frequency domain details, providing high-quality inputs that enhance detection accuracy. The framework includes two network architectures: the lightweight ASDNet, optimized for resource-limited scenarios, and SpecMFN, which combines SpecNet and MobileFaceNet for advanced feature extraction and classification. These architectures employ various anomaly scoring methods, enabling complex decision boundaries to effectively detect diverse anomalous patterns. Experimental results demonstrate that ASDNet achieves an average AUC of 94.42% and a pAUC of 87.18%, outperforming existing methods by 6.75% and 9.34%, respectively, while significantly reducing FLOPs (85.4 M, a 93.81% reduction) and parameters (0.51 M, a 41.38% reduction). SpecMFN achieves AUC and pAUC values of 94.36% and 88.60%, respectively, with FLOPs reduced by 86.6%. These results highlight the framework’s ability to balance performance and computational efficiency, making it a robust and practical solution for ASD tasks in industrial and resource-constrained environments. |
| format | Article |
| id | doaj-art-bc950d5c28ec4627a633389d6feb9469 |
| institution | Kabale University |
| issn | 2076-3417 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj-art-bc950d5c28ec4627a633389d6feb94692025-01-24T13:19:55ZengMDPI AGApplied Sciences2076-34172025-01-0115258410.3390/app15020584ASDNet: An Efficient Self-Supervised Convolutional Network for Anomalous Sound DetectionDewei Kong0Guoshun Yuan1Hongjiang Yu2Shuai Wang3Bo Zhang4Institute of Microelectronics of the Chinese Academy of Sciences, Beijing 100029, ChinaInstitute of Microelectronics of the Chinese Academy of Sciences, Beijing 100029, ChinaInstitute of Microelectronics of the Chinese Academy of Sciences, Beijing 100029, ChinaInstitute of Microelectronics of the Chinese Academy of Sciences, Beijing 100029, ChinaInstitute of Microelectronics of the Chinese Academy of Sciences, Beijing 100029, ChinaAnomalous Sound Detection (ASD) is crucial for ensuring industrial equipment safety and enhancing production efficiency. However, existing methods, while pursuing high detection accuracy, are often associated with high computational complexity, making them unsuitable for resource-constrained environments. This study proposes an efficient self-supervised ASD framework that integrates spectral features, lightweight neural networks, and various anomaly scoring methods. Unlike traditional Log-Mel features, spectral features retain richer frequency domain details, providing high-quality inputs that enhance detection accuracy. The framework includes two network architectures: the lightweight ASDNet, optimized for resource-limited scenarios, and SpecMFN, which combines SpecNet and MobileFaceNet for advanced feature extraction and classification. These architectures employ various anomaly scoring methods, enabling complex decision boundaries to effectively detect diverse anomalous patterns. Experimental results demonstrate that ASDNet achieves an average AUC of 94.42% and a pAUC of 87.18%, outperforming existing methods by 6.75% and 9.34%, respectively, while significantly reducing FLOPs (85.4 M, a 93.81% reduction) and parameters (0.51 M, a 41.38% reduction). SpecMFN achieves AUC and pAUC values of 94.36% and 88.60%, respectively, with FLOPs reduced by 86.6%. These results highlight the framework’s ability to balance performance and computational efficiency, making it a robust and practical solution for ASD tasks in industrial and resource-constrained environments.https://www.mdpi.com/2076-3417/15/2/584anomalous sound detectionlightweight neural networksASDNetSpecMFN |
| spellingShingle | Dewei Kong Guoshun Yuan Hongjiang Yu Shuai Wang Bo Zhang ASDNet: An Efficient Self-Supervised Convolutional Network for Anomalous Sound Detection Applied Sciences anomalous sound detection lightweight neural networks ASDNet SpecMFN |
| title | ASDNet: An Efficient Self-Supervised Convolutional Network for Anomalous Sound Detection |
| title_full | ASDNet: An Efficient Self-Supervised Convolutional Network for Anomalous Sound Detection |
| title_fullStr | ASDNet: An Efficient Self-Supervised Convolutional Network for Anomalous Sound Detection |
| title_full_unstemmed | ASDNet: An Efficient Self-Supervised Convolutional Network for Anomalous Sound Detection |
| title_short | ASDNet: An Efficient Self-Supervised Convolutional Network for Anomalous Sound Detection |
| title_sort | asdnet an efficient self supervised convolutional network for anomalous sound detection |
| topic | anomalous sound detection lightweight neural networks ASDNet SpecMFN |
| url | https://www.mdpi.com/2076-3417/15/2/584 |
| work_keys_str_mv | AT deweikong asdnetanefficientselfsupervisedconvolutionalnetworkforanomaloussounddetection AT guoshunyuan asdnetanefficientselfsupervisedconvolutionalnetworkforanomaloussounddetection AT hongjiangyu asdnetanefficientselfsupervisedconvolutionalnetworkforanomaloussounddetection AT shuaiwang asdnetanefficientselfsupervisedconvolutionalnetworkforanomaloussounddetection AT bozhang asdnetanefficientselfsupervisedconvolutionalnetworkforanomaloussounddetection |