A Statistical Framework for Audio Watermark Detection and Decoding
This paper introduces an integrated GMM-based blind audio watermark (WM) detection and decoding scheme that eliminates the decision threshold specification problem which constitutes drawback of the conventional decoders. The proposed method models the statistics of watermarked and original audio signals by Gaussian mixture models (GMM) with K components. Learning of the WM data is achieved in wavelet domain and a Maximum Likelihood (ML) classifier is designed for the WM decoding. Dimension of the learning space is optimized by PCA transformation. Robustness to compression, additive noise and the Stirmark benchmark attacks has been evaluated. It is shown that both WM decoding and detection performance of the introduced integrated scheme outperforms conventional correlation-based decoders. Test results demonstrate that learning in the wavelet domain improves robustness to attacks while reducing complexity. Although performance of the proposed GMM-modeling is slightly better than the SVM-based decoder introduced in , significant decrease in computational complexity makes the new method appealing.
KeywordsSupport Vector Machine Gaussian Mixture Model Wavelet Domain Detail Coefficient Training Vector
Unable to display preview. Download preview PDF.
- 1.Kirbiz, S., Gunsel, B.: Perceptual Audio Watermarking by Learning in Wavelet Domain. In: Proc. of ICPR, Hong Kong (to appear, 2006)Google Scholar
- 3.Yaslan, Y., Gunsel, B.: An Integrated Decoding Framework for Audio Watermark Extraction. In: Proc. of the ICPR 2004, Cambridge, UK, vol. 2, pp. 879–882 (August 2004)Google Scholar
- 4.Chen, T.P.C., Chen, T.: A framework for optimal blind watermark detection. In: Proc. of ACM Workshop on Multimedia and Security: New Challenges (2001)Google Scholar
- 5.Kirbiz, S., Gunsel, B.: Robust Audio Watermark Decoding by Supervised Learning. In: Proc. of IEEE ICASSP, France (May 2006)Google Scholar