Audio Watermarking Based on Partial MP3 Encoding

刘伟,王朔中,张新鹏
DOI: https://doi.org/10.3321/j.issn:0529-6579.2004.z2.007
2004-01-01
Abstract:An audio watermarking scheme based on partial MP3 encoding is proposed. The original PCM signal is first undergone a partial MP3 encoding to produce quantized MDCT coefficients. An appropriate band in the count1 region, in which magnitude of any quantized MDCT coefficient is either 1 or 0, is chosen for data embedding. Each pair of the quantized MDCT coefficients is used to hide 1 bit. After modification of the coefficients, partial decoder is performed to return to the PCM domain. As the algorithm has taken account the characteristics of MP3 encoding, it is robust against MP3. The method also has good inaudibility and large embedding capacity.
What problem does this paper attempt to address?