Anomalous Sound Detection for Industrial Machines Using Temporal Attention and Contrastive Learning

Abstract

Research in Anomalous Sound Detection (ASD) in industrial machines is a vital area of research aimed at improving operational efficiency, ensuring safety, and optimizing maintenance practices. This study leverages advanced machine learning techniques, specifically temporal attention and contrastive learning, to detect anomalies in the ASD systems. Existing approaches often face challenges such as high false positive rates, poor performance in noisy environments, and difficulties adapting to dynamic changes in machine sound patterns over time