Audio Watermarking through Deterministic plus Stochastic Signal Decomposition

 This paper describes an audio watermarking scheme based on sinusoidal signal modeling. To embed a watermark in an original signal (referred to as a cover signal hereafter), the following steps are taken. (a) A short-time Fourier transform is applied to the cover signal. (b) Prominent spectral peaks are identified and removed. (c) Their frequencies are subjected to quantization index modulation. (d) Quantized spectral peaks are added back to the spectrum. (e) Inverse Fourier transform and overlap-adding produce a watermarked signal. To decode the watermark, frequencies of prominent spectral peaks are estimated by quadratic interpolation on themagnitude spectrum. Afterwards, a maximum-likelihood procedure determines the binary value embedded in each frame. Results of testing against lossy compression, low- and highpass filtering, reverberation, and stereo-to-mono reduction are reported. A Hamming code is adopted to reduce the bit error rate (BER), and ways to improve sound quality are suggested as future research directions.