Noise Supervised Contrastive Learning and Feature-Perturbed for Anomalous Sound Detection
This work addresses false alarms in anomalous sound detection for industrial monitoring, representing a strong incremental improvement over existing methods.
The paper tackled the problem of frequent false alarms in unsupervised anomalous sound detection by introducing a one-stage supervised contrastive learning technique with feature perturbation, achieving up to 95.71% AUC, 90.23% pAUC, and 91.23% mAUC on the DCASE 2020 Challenge Task 2.
Unsupervised anomalous sound detection aims to detect unknown anomalous sounds by training a model using only normal audio data. Despite advancements in self-supervised methods, the issue of frequent false alarms when handling samples of the same type from different machines remains unresolved. This paper introduces a novel training technique called one-stage supervised contrastive learning (OS-SCL), which significantly addresses this problem by perturbing features in the embedding space and employing a one-stage noisy supervised contrastive learning approach. On the DCASE 2020 Challenge Task 2, it achieved 94.64\% AUC, 88.42\% pAUC, and 89.24\% mAUC using only Log-Mel features. Additionally, a time-frequency feature named TFgram is proposed, which is extracted from raw audio. This feature effectively captures critical information for anomalous sound detection, ultimately achieving 95.71\% AUC, 90.23\% pAUC, and 91.23\% mAUC. The source code is available at: \underline{www.github.com/huangswt/OS-SCL}.