This paper presents a local audio watermarking method based on the dual-tree complex wavelet transform (DT CWT) and distortion-compensated dither modulation (DC-DM). Specifically, we perform DT CWT on extracted audio segment, and embed the watermark signal in the decomposed low-pass coefficients. During the embedding process, the selected coefficients are block-divided into multiple host vectors, and the watermark signal is embedded into the selected set of host vectors by the dither modulation process. During the extraction process, the DC-DM is applied to generate the statistical differences, and the minimum distance decoding is employed to extract the watermark signal. Experimental results show that the proposed method is robust against common signal processing attacks and de-synchronization attacks, such as re-quantization, resampling, MP3 compression, cropping, and so on. Comparisons with the existing methods also show the superiority of our proposed method.
- Distortion-compensated dither modulation (DC-DM)
- Dual-tree complex wavelet transform (DT CWT)
- local audio watermarking
- minimum distance decoding