M. Gulbis, E. Mller, and M. Steinebach (Germany)
audio authentication watermarking, content feature point synchronization
In the research field of audio watermarking the issue of desynchronization between embedder and detector is still a challenge. For the single random cropping attack syn chronization techniques for robust watermarking were in troduced. In [1] we presented a content fragile watermark ing approach. This approach deals with desynchroniza tion caused by addition, cutting or rearrangement of larger groups of samples. Known synchronization techniques de signed for single random errors could not meet our require ments. In this paper we present a synchronization method using feature point synchronization combined with a syn chronization mark. We create a synchronization block al lowing a fast and sample-precise synchronization. Our method is verified using audio speech data with an over all playback time of about five hours.
Important Links:
Go Back
