Quantitative Analysis of Proxy Tasks for Anomalous Sound Detection

TL;DR

This paper systematically investigates how different proxy tasks impact anomalous sound detection performance, revealing that only source separation consistently correlates with improved ASD, and proposes a protocol for designing effective proxy tasks.

Contribution

It provides a quantitative analysis of proxy task effectiveness for ASD, highlighting the importance of task difficulty and objective alignment, and introduces a three-stage verification protocol.

Findings

Source separation correlates strongly with ASD performance.

Classification tasks show performance saturation due to insufficient difficulty.

Contrastive learning fails to learn meaningful features with limited data diversity.

Abstract

Anomalous sound detection (ASD) typically involves self-supervised proxy tasks to learn feature representations from normal sound data, owing to the scarcity of anomalous samples. In ASD research, proxy tasks such as AutoEncoders operate under the explicit assumption that models trained on normal data will increase the reconstruction errors related to anomalies. A natural extension suggests that improved proxy task performance should improve ASD capability; however, this relationship has received little systematic attention. This study addresses this research gap by quantitatively analyzing the relationship between proxy task metrics and ASD performance across five configurations, namely, AutoEncoders, classification, source separation, contrastive learning, and pre-trained models. We evaluate the learned representations using linear probe (linear separability) and Mahalanobis distance (distributional compactness). Our experiments reveal that strong proxy performance does not necessarily improve anomalous sound detection performance. Specifically, classification tasks experience performance saturation owing to insufficient task difficulty, whereas contrastive learning fails to learn meaningful features owing to limited data diversity. Notably, source separation is the only task demonstrating a strong positive correlation, such that improved separation consistently improves anomaly detection. Based on these findings, we highlight the critical importance of task difficulty and objective alignment. Finally, we propose a three-stage alignment verification protocol to guide the design of highly effective proxy tasks for ASD systems.