Speech Watermarking Method Using McAdams Coefficient Based on Random Forest Learning.

Candy Olivia Mawalim,Masashi Unoki

doi:10.3390/e23101246

Abstract

Speech watermarking has become a promising solution for protecting the security of speech communication systems. We propose a speech watermarking method that uses the McAdams coefficient, which is commonly used for frequency harmonics adjustment. The embedding process was conducted, using bit-inverse shifting. We also developed a random forest classifier, using features related to frequency harmonics for blind detection. An objective evaluation was conducted to analyze the performance of our method in terms of the inaudibility and robustness requirements. The results indicate that our method satisfies the speech watermarking requirements with a 16 bps payload under normal conditions and numerous non-malicious signal processing operations, e.g., conversion to Ogg or MP4 format.

Highlights

Speech communication technology has greatly advanced, due to its application in daily life
Line spectral frequencies (LSFs)), we hypothesize that this coefficient is suitable for speech watermarking. Another novelty present in this study is that we propose a speech watermarking method based on a machine learning model
The results indicate a similar tendency for all these metrics when using a larger gap between α0 and α1, i.e., better detectability, except a slight anomaly in false acceptance rate (FAR) for payloads of 16 and 32 bps

Summary

Introduction

Speech communication technology has greatly advanced, due to its application in daily life. This technology is usually implemented via a communication channel, such as the public switched telephone network (PSTN) and voice over internet protocol (VoIP). The speech communication channel is considerably vulnerable to attacks; protection and prevention countermeasures are indispensable in speech research. The recent technologies for voice conversion and text-to-speech systems are capable of using speech tampering or spoofing [1,2]

Methods

Results

Conclusion