Blind audio watermarking using adaptive quantization and Lifting wavelet transform

This work proposes an efficient and fast watermarking scheme based on lifting wavelet transform (LWT) and quantization. The approximation coefficients of LWT are firstly grouped and then their sum in a group is quantized using a quantization parameter, determined adaptively. The watermark is then extracted by comparing the sum of coefficients in a group and does not require original audio signal. Experimental analysis show higher payload capacity without compromising quality and robustness.