Exploring possible methods for Audio Anomaly Detection - on machine sounds (MIMII dataset)

Malfunctioning Industrial Machine Investigation and Inspection

🎵 A PyTorch implementation of a self-supervised complex neural network for machine sound anomaly detection using phase-aware spectral analysis and attention mechanisms.

masd: machinery anomalous sound detection

This project implements a pipeline for preprocessing, training, and evaluating Wasserstein Autoencoders (WAE) on the MIMII dataset. The pipeline includes data preprocessing, cycle generation, model training, and evaluation. Table of Contents

The repository is about sound based health monitoring of the industrial machines using machine learning.