Unified AI for Accurate Audio Anomaly Detection

TL;DR

This paper introduces a comprehensive AI framework that combines noise reduction, feature extraction, and diverse modeling techniques to improve audio anomaly detection accuracy in noisy and real-time environments.

Contribution

It presents a unified approach integrating traditional and deep learning methods for enhanced audio anomaly detection performance.

Findings

Superior accuracy on benchmark datasets

Effective noise reduction and feature extraction pipeline

Robust detection in noisy and real-time scenarios

Abstract

This paper presents a unified AI framework for high-accuracy audio anomaly detection by integrating advanced noise reduction, feature extraction, and machine learning modeling techniques. The approach combines spectral subtraction and adaptive filtering to enhance audio quality, followed by feature extraction using traditional methods like MFCCs and deep embeddings from pre-trained models such as OpenL3. The modeling pipeline incorporates classical models (SVM, Random Forest), deep learning architectures (CNNs), and ensemble methods to boost robustness and accuracy. Evaluated on benchmark datasets including TORGO and LibriSpeech, the proposed framework demonstrates superior performance in precision, recall, and classification of slurred vs. normal speech. This work addresses challenges in noisy environments and real-time applications and provides a scalable solution for audio-based anomaly detection.