Low-complexity Attention-based Unsupervised Anomalous Sound Detection exploiting Separable Convolutions and Angular Loss

TL;DR

This paper introduces a low-complexity neural network using attention and separable convolutions for unsupervised sound anomaly detection, achieving high accuracy with fewer parameters on the DCASE 2020 dataset.

Contribution

The paper proposes a novel neural network architecture that combines attention modules and separable convolutions to improve efficiency and accuracy in unsupervised sound anomaly detection.

Findings

Outperforms state-of-the-art methods in accuracy

Uses fewer parameters for comparable or better performance

Validated on DCASE 2020 challenge dataset

Abstract

In this work, a novel deep neural network, designed to enhance the efficiency and effectiveness of unsupervised sound anomaly detection, is presented. The proposed model exploits an attention module and separable convolutions to identify salient time-frequency patterns in audio data to discriminate between normal and anomalous sounds with reduced computational complexity. The approach is validated through extensive experiments using the Task 2 dataset of the DCASE 2020 challenge. Results demonstrate superior performance in terms of anomaly detection accuracy while having fewer parameters than state-of-the-art methods. Implementation details, code, and pre-trained models are available in https://github.com/michaelneri/unsupervised-audio-anomaly-detection.