Information hiding with maximum likelihood detector for correlated signals

In this paper, a new scaling based information hiding approach with high robustness against noise and gain attack is presented. The host signal is assumed to be stationary Gaussian with first-order autoregressive model. For data embedding, the host signal is divided into two parts, and just one patch is manipulated while the other one is kept unchanged for parameter estimation. A maximum likelihood (ML) decoder is proposed which uses the ratio of samples for decoding the watermarked data. Due to the decorrelating property of the proposed decoder, it is very efficient for watermarking highly correlated signals for which the decoding process is not straightforward. By calculating the distribution of the decision variable, the performance of the decoder is analytically studied. To verify the validity of the proposed algorithm, it is applied to artificial Gaussian autoregressive signals. Simulation results for highly correlated host signals confirm the robustness of our decoder.