arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2409.05035cs.SDcs.AIeess.AS
收藏

用于域泛化异常声音检测的深度通用表征

Deep Generic Representations for Domain-Generalized Anomalous Sound Detection

Tokyo Institute of Technology(东京工业大学)

Phurich Saengthong, Takahiro Shinozaki

AI总结:

针对域迁移和标注数据受限下的异常声音检测,提出 GenRep,利用大规模预训练通用表征、kNN、MemMixup 与域归一化,在 DCASE2023T2 Eval 上以 73.79% 官方分数超过最佳 OE 方法。

摘要:

开发可靠的异常声音检测(ASD)系统需要对噪声具有鲁棒性、能够适应域迁移,并在训练数据有限时保持有效性能。当前领先方法依赖每种目标机器类型的大量标注数据,利用离群点暴露(OE)技术训练特征提取器,但其在目标域上的性能仍不理想。本文提出 GenRep,它使用来自鲁棒的大规模预训练特征提取器的通用特征表征,并结合 kNN 实现域泛化 ASD,无需微调。GenRep 引入 MemMixup,这是一种利用最近源样本增强目标记忆库的简单方法,并配合域归一化技术解决源域与目标域之间的不平衡问题。在无需标注数据的情况下,GenRep 在 DCASE2023T2 Eval 集上取得 73.79% 的官方分数,优于最佳的基于 OE 的方法,并展现出有限数据场景下的鲁棒性。代码已开源。

英文摘要:

Developing a reliable anomalous sound detection (ASD) system requires robustness to noise, adaptation to domain shifts, and effective performance with limited training data. Current leading methods rely on extensive labeled data for each target machine type to train feature extractors using Outlier-Exposure (OE) techniques, yet their performance on the target domain remains sub-optimal. In this paper, we present \textit{GenRep}, which utilizes generic feature representations from a robust, large-scale pre-trained feature extractor combined with kNN for domain-generalized ASD, without the need for fine-tuning. \textit{GenRep} incorporates MemMixup, a simple approach for augmenting the target memory bank using nearest source samples, paired with a domain normalization technique to address the imbalance between source and target domains. \textit{GenRep} outperforms the best OE-based approach without a need for labeled data with an Official Score of 73.79\% on the DCASE2023T2 Eval set and demonstrates robustness under limited data scenarios. The code is available open-source.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑