Exploring Self-Supervised Audio Models for Generalized Anomalous Sound Detection

TL;DR

This paper presents a robust self-supervised audio model for generalized anomalous sound detection, leveraging pre-trained models, LoRA fine-tuning, and a machine-aware adapter to improve performance across diverse datasets.

Contribution

It introduces a novel ASD framework combining self-supervised pre-training, LoRA adaptation, and a machine-aware adapter to enhance generalization and effectiveness in diverse acoustic environments.

Findings

Significant performance improvements on all benchmark datasets.

Effective mitigation of overfitting with LoRA fine-tuning.

Enhanced generalization through machine-aware adapters.

Abstract

Machine anomalous sound detection (ASD) is a valuable technique across various applications. However, its generalization performance is often limited due to challenges in data collection and the complexity of acoustic environments. Inspired by the success of large pre-trained models in numerous fields, this paper introduces a robust ASD model that leverages self-supervised pre-trained models trained on large-scale speech and audio datasets. Although there are inconsistencies between the pre-training datasets and the ASD task, our findings indicate that pre-training still provides substantial benefits for ASD. To mitigate overfitting and retain learned knowledge when fine-tuning with limited data, we explore Fully-Connected Low-Rank Adaptation (LoRA) as an alternative to full fine-tuning. Additionally, we propose a Machine-aware Group Adapter module, which enables the model to capture differences between various machines within a unified framework, thereby enhancing the generalization performance of ASD systems. To address the challenge of missing attribute labels, we design a novel objective function that dynamically clusters unattributed data using vector quantization and optimizes through a dual-level contrastive learning loss. The proposed methods are evaluated on all benchmark datasets, including the DCASE 2020-2024 five ASD challenges, and the experimental results show significant improvements of our new approach and demonstrate the effectiveness of our proposed strategies.