• DocumentCode
    2154897
  • Title

    Detection-theoretic analysis of desynchronization attacks in watermarking

  • Author

    Moulin, Pierre ; Briassouli, Alexia ; Malvar, Henrique

  • Author_Institution
    Beckman Inst. for Adv. Sci. & Technol., Illinois Univ., Urbana, IL, USA
  • Volume
    1
  • fYear
    2002
  • fDate
    2002
  • Firstpage
    77
  • Abstract
    This paper studies the effects of desynchronization attacks such as delay and warping on the performance of blind watermark detection systems. Detection performance is measured using a probability-of-error criterion. Evaluation of the optimal likelihood ratio test is often computationally expensive, so as a practical alternative, we propose a family of quadratic detectors and construct the detector and family of watermarks that maximize the deflection criterion. For delay attacks, the deflection criterion is shown to increase quadratically with the duration of the host signal. For warping attacks, the deflection criterion increases proportionally to the duration of the signal and proportionally to the correlation time of the warping function. Piecewise-sinusoidal watermarks are shown to be nearly optimal in that case. Some examples are presented.
  • Keywords
    delays; signal detection; synchronisation; watermarking; blind watermark detection systems; correlation time; deflection criterion; delay; desynchronization attacks; detection performance; detection-theoretic analysis; optimal likelihood ratio test; piecewise-sinusoidal watermarks; probability-of-error criterion; quadratic detectors; signal duration; warping; Bayesian methods; Delay effects; Density functional theory; Detectors; Gaussian noise; Light rail systems; Probability density function; Q factor; Testing; Watermarking;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Digital Signal Processing, 2002. DSP 2002. 2002 14th International Conference on
  • Print_ISBN
    0-7803-7503-3
  • Type

    conf

  • DOI
    10.1109/ICDSP.2002.1027820
  • Filename
    1027820