Back to Search
Start Over
DNN-based speech watermarking resistant to desynchronization attacks.
- Source :
- International Journal of Wavelets, Multiresolution & Information Processing; Sep2023, Vol. 21 Issue 5, p1-20, 20p
- Publication Year :
- 2023
-
Abstract
- Desynchronization attacks proved to be the greatest challenge to audio watermarking systems as they introduce misalignment between the signal carrier and the watermark. This paper proposes a DNN-based speech watermarking system with two adversarial networks jointly trained on a set of desynchronization attacks to embed a randomly generated watermark. The detector neural network is expanded with spatial pyramid pooling layers to be able to handle signals affected by these attacks. A detailed training procedure of the aforementioned DNN system with gradual attack introduction is proposed in order to achieve robustness. Experiments performed on a speech dataset show that the system achieves satisfactory results according to all the benchmarks it was tested against. The system preserves signal quality after watermark embedding. Most importantly, the system achieved resistance to all considered desynchronization attacks. The majority of the attacks cause less than 1. 7 0 % of incorrectly detected watermarked bits on average, which outperforms comparative techniques in this regard. [ABSTRACT FROM AUTHOR]
- Subjects :
- DIGITAL watermarking
WATERMARKS
ARTIFICIAL neural networks
SOUND systems
Subjects
Details
- Language :
- English
- ISSN :
- 02196913
- Volume :
- 21
- Issue :
- 5
- Database :
- Complementary Index
- Journal :
- International Journal of Wavelets, Multiresolution & Information Processing
- Publication Type :
- Academic Journal
- Accession number :
- 169850120
- Full Text :
- https://doi.org/10.1142/S0219691323500091