An anomalous sound detection methodology for predictive maintenance

作者：

Highlights：

• Machine audio signal has been analyzed by our methodology for anomaly detection task.

• Flexibility of our methodologies allows to integrate any type of autoencoder.

• Our approach has been evaluated on real dataset in terms of efficiency and efficacy.

• We evaluate our methodology in real-time scenario having been tested on real dataset.

摘要

•Machine audio signal has been analyzed by our methodology for anomaly detection task.•Flexibility of our methodologies allows to integrate any type of autoencoder.•Our approach has been evaluated on real dataset in terms of efficiency and efficacy.•We evaluate our methodology in real-time scenario having been tested on real dataset.

论文关键词：Anomalous Sound Detection,Deep learning,Predictive maintenance

论文评审过程：Received 29 October 2021, Revised 9 July 2022, Accepted 28 July 2022, Available online 2 August 2022, Version of Record 9 August 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.118324