Audio Cross Verification Using Dual Alignment Likelihood Ratio Test

2026-07-20

AI Summary

This study addresses the challenge of detecting audio manipulations in short videos derived from news recordings by proposing an external consistency verification framework based on dual-alignment likelihood ratio testing. The method aligns the questioned audio with a trusted reference recording in an optimal manner and formulates two competing hypotheses—unaltered versus tampered—computing their likelihood ratio to assess audio consistency. Compared to conventional approaches relying on MFCC features and Euclidean distance, the proposed framework demonstrates significant improvements in computational efficiency, robustness, and interpretability. It achieves superior performance on real-world news audio verification tasks, offering a novel and effective direction for audio forensic analysis.

Abstract

This paper explores a way to verify that audio has not been maliciously tampered in a specific context: short viral videos taken from news recordings. Rather than trying to detect artifacts of tampering (internal inconsistency), we focus on positively verifying a query against a trusted source such as a news recording (external consistency). We propose a method for cross verifying a short audio query against a reference recording from which it was taken. Our approach is to define two hypotheses (non-tampered vs tampered), calculate the most likely alignment between query and reference for each hypothesis, and then perform a likelihood ratio test on the two alignments. We show that this method is fast to compute, much more robust than using MFCC features with Euclidean distance, and has the key benefit of explainability. Our cross verification approach provides an alternative perspective and complementary tool to existing tampering detection methods.

Problem

• audio verification
• tampering detection
• external consistency
• short viral videos
• trusted source

Innovation

• audio cross verification
• dual alignment
• likelihood ratio test
• external consistency
• tampering detection