Unsupervised anomalous sound detection (ASD) aims to identify anomalous
sounds by learning the features of normal operational sounds and sensing their
deviations. Recent approaches have focused on the self-supervised task
utilizing the classification of normal data, and advanced models have shown
that securing representation space for anomalous data is important through
representation learning yielding compact intra-class and well-separated
intra-class distributions. However, we show that conventional approaches often
fail to ensure sufficient intra-class compactness and exhibit angular disparity
between samples and their corresponding centers. In this paper, we propose a
training technique aimed at ensuring intra-class compactness and increasing the
angle gap between normal and abnormal samples. Furthermore, we present an
architecture that extracts features for important temporal regions, enabling
the model to learn which time frames should be emphasized or suppressed.
Experimental results demonstrate that the proposed method achieves the best
performance giving 0.90%, 0.83%, and 2.16% improvement in terms of AUC, pAUC,
and mAUC, respectively, compared to the state-of-the-art method on DCASE 2020
Challenge Task2 dataset.

通过学习正常操作声音的特征并感知其偏差，无监督异常声音检测（ASD）旨在识别异常声音。近期研究集中于自监督任务，利用正常数据的分类，而先进模型的研究表明，通过表示学习及产生紧凑的类内和互不重叠的类间分布，保证异常数据的表示空间是重要的。然而，传统方法通常无法确保足够的类内紧凑性，并存在样本与其相应中心之间的角差异。本文提出了一种训练技术，旨在确保类内紧凑性并增加正常和异常样本之间的角间隔。此外，我们提出了一种提取重要时间区域特征的结构，使模型能够学习哪些时间帧应该被强调或抑制。实验结果表明，该方法在 DCASE 2020 Challenge Task2 数据集上相对于最先进的方法，AUC、pAUC 和 mAUC 分别提高了 0.90％、0.83％和 2.16％。