Learning ancestral atom via sparse coding

Toshimitsu Aritake, Hideitsu Hino, Noboru Murata

    Research output: Contribution to journalArticle

    2 Citations (Scopus)

    Abstract

    Sparse signal models have been the focus of recent research. In sparse coding, signals are represented with a linear combination of a small number of elementary signals called atoms, and the collection of atoms is called a dictionary. Design of the dictionary has strong influence on the signal approximation performance. Recently, to put prior information into dictionary learning, several methods imposing a certain kind of structure on the dictionary are proposed. In this paper, like wavelet analysis, a dictionary for sparse signal representation is assumed to be generated from an ancestral atom, and a method for learning the ancestral atom is proposed. The proposed algorithm updates the ancestral atom by iterating dictionary update in unstructured dictionary space and projection of the updated dictionary onto the structured dictionary space. The algorithm allows a simple differential geometric interpretation. Numerical experiments are performed to show the characteristics and advantages of the proposed algorithm.

    Original languageEnglish
    Article number6412707
    Pages (from-to)586-594
    Number of pages9
    JournalIEEE Journal on Selected Topics in Signal Processing
    Volume7
    Issue number4
    DOIs
    Publication statusPublished - 2013

    Fingerprint

    Glossaries
    Atoms
    Wavelet analysis

    Keywords

    • Atom decomposition
    • sparse representation
    • structured dictionary learning

    ASJC Scopus subject areas

    • Electrical and Electronic Engineering
    • Signal Processing

    Cite this

    Learning ancestral atom via sparse coding. / Aritake, Toshimitsu; Hino, Hideitsu; Murata, Noboru.

    In: IEEE Journal on Selected Topics in Signal Processing, Vol. 7, No. 4, 6412707, 2013, p. 586-594.

    Research output: Contribution to journalArticle

    Aritake, Toshimitsu ; Hino, Hideitsu ; Murata, Noboru. / Learning ancestral atom via sparse coding. In: IEEE Journal on Selected Topics in Signal Processing. 2013 ; Vol. 7, No. 4. pp. 586-594.
    @article{733af725b1dc452db2419abc2424dea6,
    title = "Learning ancestral atom via sparse coding",
    abstract = "Sparse signal models have been the focus of recent research. In sparse coding, signals are represented with a linear combination of a small number of elementary signals called atoms, and the collection of atoms is called a dictionary. Design of the dictionary has strong influence on the signal approximation performance. Recently, to put prior information into dictionary learning, several methods imposing a certain kind of structure on the dictionary are proposed. In this paper, like wavelet analysis, a dictionary for sparse signal representation is assumed to be generated from an ancestral atom, and a method for learning the ancestral atom is proposed. The proposed algorithm updates the ancestral atom by iterating dictionary update in unstructured dictionary space and projection of the updated dictionary onto the structured dictionary space. The algorithm allows a simple differential geometric interpretation. Numerical experiments are performed to show the characteristics and advantages of the proposed algorithm.",
    keywords = "Atom decomposition, sparse representation, structured dictionary learning",
    author = "Toshimitsu Aritake and Hideitsu Hino and Noboru Murata",
    year = "2013",
    doi = "10.1109/JSTSP.2013.2240254",
    language = "English",
    volume = "7",
    pages = "586--594",
    journal = "IEEE Journal on Selected Topics in Signal Processing",
    issn = "1932-4553",
    publisher = "Institute of Electrical and Electronics Engineers Inc.",
    number = "4",

    }

    TY - JOUR

    T1 - Learning ancestral atom via sparse coding

    AU - Aritake, Toshimitsu

    AU - Hino, Hideitsu

    AU - Murata, Noboru

    PY - 2013

    Y1 - 2013

    N2 - Sparse signal models have been the focus of recent research. In sparse coding, signals are represented with a linear combination of a small number of elementary signals called atoms, and the collection of atoms is called a dictionary. Design of the dictionary has strong influence on the signal approximation performance. Recently, to put prior information into dictionary learning, several methods imposing a certain kind of structure on the dictionary are proposed. In this paper, like wavelet analysis, a dictionary for sparse signal representation is assumed to be generated from an ancestral atom, and a method for learning the ancestral atom is proposed. The proposed algorithm updates the ancestral atom by iterating dictionary update in unstructured dictionary space and projection of the updated dictionary onto the structured dictionary space. The algorithm allows a simple differential geometric interpretation. Numerical experiments are performed to show the characteristics and advantages of the proposed algorithm.

    AB - Sparse signal models have been the focus of recent research. In sparse coding, signals are represented with a linear combination of a small number of elementary signals called atoms, and the collection of atoms is called a dictionary. Design of the dictionary has strong influence on the signal approximation performance. Recently, to put prior information into dictionary learning, several methods imposing a certain kind of structure on the dictionary are proposed. In this paper, like wavelet analysis, a dictionary for sparse signal representation is assumed to be generated from an ancestral atom, and a method for learning the ancestral atom is proposed. The proposed algorithm updates the ancestral atom by iterating dictionary update in unstructured dictionary space and projection of the updated dictionary onto the structured dictionary space. The algorithm allows a simple differential geometric interpretation. Numerical experiments are performed to show the characteristics and advantages of the proposed algorithm.

    KW - Atom decomposition

    KW - sparse representation

    KW - structured dictionary learning

    UR - http://www.scopus.com/inward/record.url?scp=84880912802&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84880912802&partnerID=8YFLogxK

    U2 - 10.1109/JSTSP.2013.2240254

    DO - 10.1109/JSTSP.2013.2240254

    M3 - Article

    AN - SCOPUS:84880912802

    VL - 7

    SP - 586

    EP - 594

    JO - IEEE Journal on Selected Topics in Signal Processing

    JF - IEEE Journal on Selected Topics in Signal Processing

    SN - 1932-4553

    IS - 4

    M1 - 6412707

    ER -