Extension of Hidden Markov models for multiple candidates and its application to gesture recognition

    Research output: Contribution to journalArticle

    1 Citation (Scopus)

    Abstract

    We propose a modified Hidden Markov Model (HMM) with a view to improve gesture recognition using a moving camera. The conventional HMM is formulated so as to deal with only one feature candidate per frame. However, for a mobile robot, the background and the lighting conditions are always changing, and the feature extraction problem becomes difficult. It is almost impossible to extract a reliable feature vector under such conditions. In this paper, we define a new gesture recognition framework in which multiple candidates of feature vectors are generated with confidence measures and the HMM is extended to deal with these multiple feature vectors. Experimental results comparing the proposed system with feature vectors based on DCT and the method of selecting only one candidate feature point verifies the effectiveness of the proposed technique.

    Original languageEnglish
    Pages (from-to)1239-1246
    Number of pages8
    JournalIEICE Transactions on Information and Systems
    VolumeE88-D
    Issue number6
    DOIs
    Publication statusPublished - 2005 Jun

    Fingerprint

    Gesture recognition
    Hidden Markov models
    Mobile robots
    Feature extraction
    Lighting
    Cameras

    Keywords

    • Gesture recognition
    • Hidden Markov Model
    • Mobile robot
    • Multiple candidates of feature vector

    ASJC Scopus subject areas

    • Information Systems
    • Computer Graphics and Computer-Aided Design
    • Software

    Cite this

    @article{3e364070e5814a6b86eac40b3fd6549d,
    title = "Extension of Hidden Markov models for multiple candidates and its application to gesture recognition",
    abstract = "We propose a modified Hidden Markov Model (HMM) with a view to improve gesture recognition using a moving camera. The conventional HMM is formulated so as to deal with only one feature candidate per frame. However, for a mobile robot, the background and the lighting conditions are always changing, and the feature extraction problem becomes difficult. It is almost impossible to extract a reliable feature vector under such conditions. In this paper, we define a new gesture recognition framework in which multiple candidates of feature vectors are generated with confidence measures and the HMM is extended to deal with these multiple feature vectors. Experimental results comparing the proposed system with feature vectors based on DCT and the method of selecting only one candidate feature point verifies the effectiveness of the proposed technique.",
    keywords = "Gesture recognition, Hidden Markov Model, Mobile robot, Multiple candidates of feature vector",
    author = "Yosuke Sato and Tetsuji Ogawa and Tetsunori Kobayashi",
    year = "2005",
    month = "6",
    doi = "10.1093/ietisy/e88-d.6.1239",
    language = "English",
    volume = "E88-D",
    pages = "1239--1246",
    journal = "IEICE Transactions on Information and Systems",
    issn = "0916-8532",
    publisher = "Maruzen Co., Ltd/Maruzen Kabushikikaisha",
    number = "6",

    }

    TY - JOUR

    T1 - Extension of Hidden Markov models for multiple candidates and its application to gesture recognition

    AU - Sato, Yosuke

    AU - Ogawa, Tetsuji

    AU - Kobayashi, Tetsunori

    PY - 2005/6

    Y1 - 2005/6

    N2 - We propose a modified Hidden Markov Model (HMM) with a view to improve gesture recognition using a moving camera. The conventional HMM is formulated so as to deal with only one feature candidate per frame. However, for a mobile robot, the background and the lighting conditions are always changing, and the feature extraction problem becomes difficult. It is almost impossible to extract a reliable feature vector under such conditions. In this paper, we define a new gesture recognition framework in which multiple candidates of feature vectors are generated with confidence measures and the HMM is extended to deal with these multiple feature vectors. Experimental results comparing the proposed system with feature vectors based on DCT and the method of selecting only one candidate feature point verifies the effectiveness of the proposed technique.

    AB - We propose a modified Hidden Markov Model (HMM) with a view to improve gesture recognition using a moving camera. The conventional HMM is formulated so as to deal with only one feature candidate per frame. However, for a mobile robot, the background and the lighting conditions are always changing, and the feature extraction problem becomes difficult. It is almost impossible to extract a reliable feature vector under such conditions. In this paper, we define a new gesture recognition framework in which multiple candidates of feature vectors are generated with confidence measures and the HMM is extended to deal with these multiple feature vectors. Experimental results comparing the proposed system with feature vectors based on DCT and the method of selecting only one candidate feature point verifies the effectiveness of the proposed technique.

    KW - Gesture recognition

    KW - Hidden Markov Model

    KW - Mobile robot

    KW - Multiple candidates of feature vector

    UR - http://www.scopus.com/inward/record.url?scp=25844503826&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=25844503826&partnerID=8YFLogxK

    U2 - 10.1093/ietisy/e88-d.6.1239

    DO - 10.1093/ietisy/e88-d.6.1239

    M3 - Article

    VL - E88-D

    SP - 1239

    EP - 1246

    JO - IEICE Transactions on Information and Systems

    JF - IEICE Transactions on Information and Systems

    SN - 0916-8532

    IS - 6

    ER -