• DocumentCode
    1546174
  • Title

    Block and Sliding-Block Lossy Compression via MCMC

  • Author

    Jalali, Shirin ; Weissman, Tsachy

  • Author_Institution
    Center for Mathematics of Information at California Institute of Technology
  • Volume
    60
  • Issue
    8
  • fYear
    2012
  • fDate
    8/1/2012 12:00:00 AM
  • Firstpage
    2187
  • Lastpage
    2198
  • Abstract
    We propose an approach to lossy compression of finite-alphabet sources that utilizes Markov chain Monte Carlo (MCMC) and simulated annealing methods. The idea is to define an energy function over the space of reconstruction sequences. The energy of a candidate reconstruction sequence is defined such that it incorporates its distortion relative to the source sequence, its compressibility, and the point sought on the rate-distortion curve. The proposed algorithm samples from the Boltzmann distribution associated with this energy function using the "heat-bath" algorithm. The complexity of each iteration is independent of the sequence length and is only linearly dependent on a certain context parameter, which grows sub-logarithmically with the sequence length. We show that the proposed algorithm achieves optimum rate-distortion performance in the limits of large number of iterations, and sequence length, when employed on any stationary ergodic source. Inspired by the proposed block-coding algorithm, we also propose an algorithm for constructing sliding-block (SB) codes using similar ideas.
  • Keywords
    Compression algorithms; Encoding; Entropy; Markov processes; Rate-distortion; Simulated annealing; Vectors; Gibbs sampler; Markov chain Monte Carlo; Rate-distortion coding; simulated annealing; universal lossy compression;
  • fLanguage
    English
  • Journal_Title
    Communications, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    0090-6778
  • Type

    jour

  • DOI
    10.1109/TCOMM.2012.061412.110194
  • Filename
    6222291