Optical fragile zero-watermarking for voice tampering detection based on single-pixel imaging and bimodal biometric keys
Applied Optics · 01.08.2026 · 65 (22) · 7396
DOI: 10.1364/ao.606552 ↗Kurz gefasst
Fachartikel in der Zeitschrift Applied Optics, online erschienen am 23.07.2026. Verfasst von Yonggang Su und 2 weiteren Autor:innen. Die Arbeit nennt 39 Literaturangaben. Gefördert von: Advanced Talents Incubation Program of Hebei University, Natural Science Foundation of Hebei Province.
Abstract (Originalsprache)
Voice is one of the most important carriers of human information communication, and its authenticity and integrity are crucial in identity authentication, forensic analysis, remote communication, and multimedia security. However, with the rapid development of audio editing and speech synthesis technologies, voice signals are increasingly exposed to tampering, forgery, and unauthorized replacement. Existing optical robust and fragile watermarking methods are mainly designed for image content protection and tampering detection, whereas optical watermarking techniques for voice tampering detection have rarely been investigated. To address this issue, this paper proposes an optical fragile zero-watermarking method for voice information tampering detection using single-pixel imaging and bimodal biometric keys. In the proposed method, bimodal biometric features are first extracted from face and palmprint images and then fused with the host-voice digest to generate voice-biometric joint chaotic initial values for driving a two-dimensional chaotic system. Subsequently, the chaotic system is used to generate random speckle illumination patterns and measurement-position indices. Based on the single-pixel measurement mechanism, the watermark image is optically encoded to obtain the encoded watermark. Meanwhile, a fragile feature derived from the time-frequency phase difference of the host voice is extracted and further combined with the encoded watermark through an XOR operation to construct the zero-watermark. When the host voice is tampered with, the extracted fragile voice feature changes, which prevents the valid watermark from being correctly recovered and enables sensitive voice tampering detection. Extensive experimental results demonstrate that the proposed method exhibits high key sensitivity and strong tampering detection capability. It can effectively distinguish authenticated voices from tampered ones while preserving the original voice signal without any modification. The proposed method provides a new perspective, to the best of our knowledge, for the integration of voice information security and optical watermarking technology.
Förderung
- Advanced Talents Incubation Program of Hebei University
- Natural Science Foundation of Hebei Province
Zitieren
APA
Su, Y., Qi, L., & Li, F. (2026). Optical fragile zero-watermarking for voice tampering detection based on single-pixel imaging and bimodal biometric keys. Applied Optics, 65(22), 7396. https://doi.org/10.1364/ao.606552
Vancouver
Su Y, Qi L, Li F. Optical fragile zero-watermarking for voice tampering detection based on single-pixel imaging and bimodal biometric keys. Applied Optics. 2026;65(22):7396. doi:10.1364/ao.606552
BibTeX
@article{Su2026,
author = {Su, Yonggang and Qi, Liang and Li, Fei},
title = {Optical fragile zero-watermarking for voice tampering detection based on single-pixel imaging and bimodal biometric keys},
journal = {Applied Optics},
year = {2026},
volume = {65},
number = {22},
pages = {7396},
doi = {10.1364/ao.606552}
}
EUVIMED