A study on backpropagation networks for parameter estimation from grey-scale images

Tian-Jin Feng, Z. Houkes, M.J. Korsten, L.J. Spreeuwers

    Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

    29 Downloads (Pure)

    Abstract

    A large number of experiments have been done on the basic research of parameter estimation from images with neural networks. To obtain a better estimation accuracy of parameters and to decrease needed storage space and computation time, the architecture of networks, the effective learning rate and momentum, and the selection of training set are investigated. A comparison of network performance to that of the least squares estimator is made. The internal representations in trained networks, i.e. input-to-hidden weight maps or measuring models, which include statistical features of training images and have a clear physical and geometrical meaning, and the internal components of output parameters given by outputs of hidden neurons are presented
    Original languageEnglish
    Title of host publicationProceedings of the 1991 IEEE International Joint Conference on Neural Networks, IJCNN'91
    Place of PublicationPiscataway, NJ
    PublisherIEEE
    Pages331-336
    Number of pages6
    Volume1
    ISBN (Print)0-7803-0227-3
    DOIs
    Publication statusPublished - 18 Nov 1991
    Event1991 IEEE International Conference on Neural Networks, ICNN 1991 - Singapore, Singapore
    Duration: 18 Nov 199121 Nov 1991

    Conference

    Conference1991 IEEE International Conference on Neural Networks, ICNN 1991
    Abbreviated titleICNN
    CountrySingapore
    CitySingapore
    Period18/11/9121/11/91

    Fingerprint

    Network performance
    Backpropagation
    Parameter estimation
    Neurons
    Momentum
    Neural networks
    Experiments

    Cite this

    Feng, T-J., Houkes, Z., Korsten, M. J., & Spreeuwers, L. J. (1991). A study on backpropagation networks for parameter estimation from grey-scale images. In Proceedings of the 1991 IEEE International Joint Conference on Neural Networks, IJCNN'91 (Vol. 1, pp. 331-336). Piscataway, NJ: IEEE. https://doi.org/10.1109/IJCNN.1991.170423
    Feng, Tian-Jin ; Houkes, Z. ; Korsten, M.J. ; Spreeuwers, L.J. / A study on backpropagation networks for parameter estimation from grey-scale images. Proceedings of the 1991 IEEE International Joint Conference on Neural Networks, IJCNN'91. Vol. 1 Piscataway, NJ : IEEE, 1991. pp. 331-336
    @inproceedings{c5b06b515009488990b381e30eac1b83,
    title = "A study on backpropagation networks for parameter estimation from grey-scale images",
    abstract = "A large number of experiments have been done on the basic research of parameter estimation from images with neural networks. To obtain a better estimation accuracy of parameters and to decrease needed storage space and computation time, the architecture of networks, the effective learning rate and momentum, and the selection of training set are investigated. A comparison of network performance to that of the least squares estimator is made. The internal representations in trained networks, i.e. input-to-hidden weight maps or measuring models, which include statistical features of training images and have a clear physical and geometrical meaning, and the internal components of output parameters given by outputs of hidden neurons are presented",
    author = "Tian-Jin Feng and Z. Houkes and M.J. Korsten and L.J. Spreeuwers",
    year = "1991",
    month = "11",
    day = "18",
    doi = "10.1109/IJCNN.1991.170423",
    language = "English",
    isbn = "0-7803-0227-3",
    volume = "1",
    pages = "331--336",
    booktitle = "Proceedings of the 1991 IEEE International Joint Conference on Neural Networks, IJCNN'91",
    publisher = "IEEE",
    address = "United States",

    }

    Feng, T-J, Houkes, Z, Korsten, MJ & Spreeuwers, LJ 1991, A study on backpropagation networks for parameter estimation from grey-scale images. in Proceedings of the 1991 IEEE International Joint Conference on Neural Networks, IJCNN'91. vol. 1, IEEE, Piscataway, NJ, pp. 331-336, 1991 IEEE International Conference on Neural Networks, ICNN 1991, Singapore, Singapore, 18/11/91. https://doi.org/10.1109/IJCNN.1991.170423

    A study on backpropagation networks for parameter estimation from grey-scale images. / Feng, Tian-Jin; Houkes, Z.; Korsten, M.J.; Spreeuwers, L.J.

    Proceedings of the 1991 IEEE International Joint Conference on Neural Networks, IJCNN'91. Vol. 1 Piscataway, NJ : IEEE, 1991. p. 331-336.

    Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

    TY - GEN

    T1 - A study on backpropagation networks for parameter estimation from grey-scale images

    AU - Feng, Tian-Jin

    AU - Houkes, Z.

    AU - Korsten, M.J.

    AU - Spreeuwers, L.J.

    PY - 1991/11/18

    Y1 - 1991/11/18

    N2 - A large number of experiments have been done on the basic research of parameter estimation from images with neural networks. To obtain a better estimation accuracy of parameters and to decrease needed storage space and computation time, the architecture of networks, the effective learning rate and momentum, and the selection of training set are investigated. A comparison of network performance to that of the least squares estimator is made. The internal representations in trained networks, i.e. input-to-hidden weight maps or measuring models, which include statistical features of training images and have a clear physical and geometrical meaning, and the internal components of output parameters given by outputs of hidden neurons are presented

    AB - A large number of experiments have been done on the basic research of parameter estimation from images with neural networks. To obtain a better estimation accuracy of parameters and to decrease needed storage space and computation time, the architecture of networks, the effective learning rate and momentum, and the selection of training set are investigated. A comparison of network performance to that of the least squares estimator is made. The internal representations in trained networks, i.e. input-to-hidden weight maps or measuring models, which include statistical features of training images and have a clear physical and geometrical meaning, and the internal components of output parameters given by outputs of hidden neurons are presented

    U2 - 10.1109/IJCNN.1991.170423

    DO - 10.1109/IJCNN.1991.170423

    M3 - Conference contribution

    SN - 0-7803-0227-3

    VL - 1

    SP - 331

    EP - 336

    BT - Proceedings of the 1991 IEEE International Joint Conference on Neural Networks, IJCNN'91

    PB - IEEE

    CY - Piscataway, NJ

    ER -

    Feng T-J, Houkes Z, Korsten MJ, Spreeuwers LJ. A study on backpropagation networks for parameter estimation from grey-scale images. In Proceedings of the 1991 IEEE International Joint Conference on Neural Networks, IJCNN'91. Vol. 1. Piscataway, NJ: IEEE. 1991. p. 331-336 https://doi.org/10.1109/IJCNN.1991.170423