Posterior mean super-resolution with a compound Gaussian Markov random field prior

Takayuki Katsuki, Masato Inoue

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    2 Citations (Scopus)

    Abstract

    This manuscript proposes a posterior mean (PM) super-resolution (SR) method with a compound Gaussian Markov random field (MRF) prior. SR is a technique to estimate a spatially high-resolution image from observed multiple low-resolution images. A compound Gaussian MRF model provides a preferable prior for natural images that preserves edges. PM is the optimal estimator for the objective function of peak signal-to-noise ratio (PSNR). This estimator is numerically determined by using variational Bayes (VB). We then solve the conjugate prior problem on VB and the exponential-order calculation cost problem of a compound Gaussian MRF prior with simple Taylor approximations. In experiments, the proposed method roughly overcomes existing methods.

    Original languageEnglish
    Title of host publicationICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
    Pages841-844
    Number of pages4
    DOIs
    Publication statusPublished - 2012
    Event2012 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2012 - Kyoto
    Duration: 2012 Mar 252012 Mar 30

    Other

    Other2012 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2012
    CityKyoto
    Period12/3/2512/3/30

    Fingerprint

    Image resolution
    Signal to noise ratio
    Costs
    Experiments

    Keywords

    • fully Bayesian approach
    • Markov random field prior
    • super-resolution
    • Taylor approximation
    • variational Bayes

    ASJC Scopus subject areas

    • Signal Processing
    • Software
    • Electrical and Electronic Engineering

    Cite this

    Katsuki, T., & Inoue, M. (2012). Posterior mean super-resolution with a compound Gaussian Markov random field prior. In ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings (pp. 841-844). [6288015] https://doi.org/10.1109/ICASSP.2012.6288015

    Posterior mean super-resolution with a compound Gaussian Markov random field prior. / Katsuki, Takayuki; Inoue, Masato.

    ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. 2012. p. 841-844 6288015.

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Katsuki, T & Inoue, M 2012, Posterior mean super-resolution with a compound Gaussian Markov random field prior. in ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings., 6288015, pp. 841-844, 2012 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2012, Kyoto, 12/3/25. https://doi.org/10.1109/ICASSP.2012.6288015
    Katsuki T, Inoue M. Posterior mean super-resolution with a compound Gaussian Markov random field prior. In ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. 2012. p. 841-844. 6288015 https://doi.org/10.1109/ICASSP.2012.6288015
    Katsuki, Takayuki ; Inoue, Masato. / Posterior mean super-resolution with a compound Gaussian Markov random field prior. ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. 2012. pp. 841-844
    @inproceedings{5e382e86e37f43159ddf64d4a9d3680e,
    title = "Posterior mean super-resolution with a compound Gaussian Markov random field prior",
    abstract = "This manuscript proposes a posterior mean (PM) super-resolution (SR) method with a compound Gaussian Markov random field (MRF) prior. SR is a technique to estimate a spatially high-resolution image from observed multiple low-resolution images. A compound Gaussian MRF model provides a preferable prior for natural images that preserves edges. PM is the optimal estimator for the objective function of peak signal-to-noise ratio (PSNR). This estimator is numerically determined by using variational Bayes (VB). We then solve the conjugate prior problem on VB and the exponential-order calculation cost problem of a compound Gaussian MRF prior with simple Taylor approximations. In experiments, the proposed method roughly overcomes existing methods.",
    keywords = "fully Bayesian approach, Markov random field prior, super-resolution, Taylor approximation, variational Bayes",
    author = "Takayuki Katsuki and Masato Inoue",
    year = "2012",
    doi = "10.1109/ICASSP.2012.6288015",
    language = "English",
    isbn = "9781467300469",
    pages = "841--844",
    booktitle = "ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings",

    }

    TY - GEN

    T1 - Posterior mean super-resolution with a compound Gaussian Markov random field prior

    AU - Katsuki, Takayuki

    AU - Inoue, Masato

    PY - 2012

    Y1 - 2012

    N2 - This manuscript proposes a posterior mean (PM) super-resolution (SR) method with a compound Gaussian Markov random field (MRF) prior. SR is a technique to estimate a spatially high-resolution image from observed multiple low-resolution images. A compound Gaussian MRF model provides a preferable prior for natural images that preserves edges. PM is the optimal estimator for the objective function of peak signal-to-noise ratio (PSNR). This estimator is numerically determined by using variational Bayes (VB). We then solve the conjugate prior problem on VB and the exponential-order calculation cost problem of a compound Gaussian MRF prior with simple Taylor approximations. In experiments, the proposed method roughly overcomes existing methods.

    AB - This manuscript proposes a posterior mean (PM) super-resolution (SR) method with a compound Gaussian Markov random field (MRF) prior. SR is a technique to estimate a spatially high-resolution image from observed multiple low-resolution images. A compound Gaussian MRF model provides a preferable prior for natural images that preserves edges. PM is the optimal estimator for the objective function of peak signal-to-noise ratio (PSNR). This estimator is numerically determined by using variational Bayes (VB). We then solve the conjugate prior problem on VB and the exponential-order calculation cost problem of a compound Gaussian MRF prior with simple Taylor approximations. In experiments, the proposed method roughly overcomes existing methods.

    KW - fully Bayesian approach

    KW - Markov random field prior

    KW - super-resolution

    KW - Taylor approximation

    KW - variational Bayes

    UR - http://www.scopus.com/inward/record.url?scp=84867584630&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84867584630&partnerID=8YFLogxK

    U2 - 10.1109/ICASSP.2012.6288015

    DO - 10.1109/ICASSP.2012.6288015

    M3 - Conference contribution

    SN - 9781467300469

    SP - 841

    EP - 844

    BT - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings

    ER -