Fig 2 - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
Source publication
We propose an acoustic anomaly detection algorithm based on the framework of contrastive learning. Contrastive learning is a recently proposed self-supervised approach that has shown promising results in image classification and speech recognition. However, its application in anomaly detection is underexplored. Earlier studies have demonstrated tha...
Similar publications
We propose an acoustic anomaly detection algorithm based on the framework of contrastive learning. Contrastive learning is a recently proposed self-supervised approach that has shown promising results in image classification and speech recognition. However, its application in anomaly detection is underexplored. Earlier studies have demonstrated tha...
We propose an acoustic anomaly detection algorithm based on the framework of contrastive learning. Contrastive learning is a recently proposed self-supervised approach that has shown promising results in image classification and speech recognition. However, its application in anomaly detection is underexplored. Earlier studies have demonstrated tha...
This study addresses the problem of single-channel Automatic Speech Recognition of a target speaker within an overlap speech scenario. In the proposed method, the hidden representations in the acoustic model are modulated by speaker auxiliary information to recognize only the desired speaker. Affine transformation layers are inserted into the acous...
Citations
Current data augmentation methods for machine anomalous sound detection (MASD) suffer from insufficient data generated by real world machines. Open datasets such as audioset are not tailored for machine sounds, and fake sounds created by generative models are not trustworthy. In this paper, we explore a novel data augmentation method in MASD using machine sounds simulated by finite element analysis (FEA). We use Ansys, a software capable for acoustic simulation based on FEA, to generate machine sounds for further training. The physical properties of the machine, such as geometry and material, and the material of the medium is modified to acquire data from multiple domains. The experimental results on DCASE 2023 Task 2 dataset indicates a better performance from models trained using augmented data.
Acoustic sensing provides crucial data for anomalous sound detection (ASD) in condition monitoring. However, building a robust acoustic-sensing-based ASD system is challenging due to the unsupervised nature of training data, which only contain normal sound samples. Recent discriminative models based on machine identity (ID) classification have shown excellent ASD performance by leveraging strong prior knowledge like machine ID. However, such strong priors are often unavailable in real-world applications, limiting these models. To address this, we propose utilizing the imbalanced and inconsistent attribute labels from acoustic sensors, such as machine running speed and microphone model, as weak priors to train an attribute classifier. We also introduce an imbalanced compensation strategy to handle extremely imbalanced categories and ensure model trainability. Furthermore, we propose a score fusion method to enhance anomaly detection robustness. The proposed algorithm was applied in our DCASE2023 Challenge Task 2 submission, ranking sixth internationally. By exploiting acoustic sensor data attributes as weak prior knowledge, our approach provides an effective framework for robust ASD when strong priors are absent.
