Detection of AI-Synthesized Speech Using Cepstral & Bispectral Statistics

TL;DR

This paper introduces a novel method combining bispectral and cepstral statistics with machine learning to effectively distinguish AI-synthesized speech from human speech, enhancing digital audio forensics.

Contribution

It presents a new approach that integrates bispectral and cepstral analysis with machine learning for improved detection of AI-generated speech.

Findings

Higher-order statistics show less correlation in human speech.

Cepstral analysis reveals a power component missing in synthesized speech.

The combined method improves detection accuracy.

Abstract

Digital technology has made possible unimaginable applications come true. It seems exciting to have a handful of tools for easy editing and manipulation, but it raises alarming concerns that can propagate as speech clones, duplicates, or maybe deep fakes. Validating the authenticity of a speech is one of the primary problems of digital audio forensics. We propose an approach to distinguish human speech from AI synthesized speech exploiting the Bi-spectral and Cepstral analysis. Higher-order statistics have less correlation for human speech in comparison to a synthesized speech. Also, Cepstral analysis revealed a durable power component in human speech that is missing for a synthesized speech. We integrate both these analyses and propose a machine learning model to detect AI synthesized speech.