Learning Video Manifolds for Content Analysis of Crowded Scenes

Myo Thida, Eng How-Lung, Dorothy Monekosso, Paolo Remagnino

    Research output: Contribution to journalArticle

    4 Citations (Scopus)

    Abstract

    In this paper, we propose a new approach for recognizing group events and abnormality detection in a crowded scene. A manifold learning algorithm with temporal-constraints is proposed to embed a video of a crowded scene in a low-dimensional space. Our low dimensional representation of a video preserves the spatial temporal property of a video as well as the characteristic of the video. Recognizing video events and abnormality detection in a crowded scene is achieved by studying the video trajectory in the manifold space. We evaluate our proposed method on the state-of-the-art public data-sets containing different crowd events. Qualitative and quantitative results show the promising performance of the proposed method.
    LanguageEnglish
    Pages71-77
    JournalIPSJ Transactions on Computer Vision and Applications
    Volume4
    DOIs
    Publication statusPublished - 30 May 2012

    Fingerprint

    Learning algorithms
    Trajectories

    Cite this

    Thida, Myo ; How-Lung, Eng ; Monekosso, Dorothy ; Remagnino, Paolo. / Learning Video Manifolds for Content Analysis of Crowded Scenes. In: IPSJ Transactions on Computer Vision and Applications. 2012 ; Vol. 4. pp. 71-77.
    @article{5de50ee12441453fb80cd387e677d392,
    title = "Learning Video Manifolds for Content Analysis of Crowded Scenes",
    abstract = "In this paper, we propose a new approach for recognizing group events and abnormality detection in a crowded scene. A manifold learning algorithm with temporal-constraints is proposed to embed a video of a crowded scene in a low-dimensional space. Our low dimensional representation of a video preserves the spatial temporal property of a video as well as the characteristic of the video. Recognizing video events and abnormality detection in a crowded scene is achieved by studying the video trajectory in the manifold space. We evaluate our proposed method on the state-of-the-art public data-sets containing different crowd events. Qualitative and quantitative results show the promising performance of the proposed method.",
    author = "Myo Thida and Eng How-Lung and Dorothy Monekosso and Paolo Remagnino",
    year = "2012",
    month = "5",
    day = "30",
    doi = "10.2197/ipsjtcva.4.71",
    language = "English",
    volume = "4",
    pages = "71--77",
    journal = "IPSJ Transactions on Computer Vision and Applications",
    issn = "1882-6695",

    }

    Learning Video Manifolds for Content Analysis of Crowded Scenes. / Thida, Myo; How-Lung, Eng; Monekosso, Dorothy; Remagnino, Paolo.

    In: IPSJ Transactions on Computer Vision and Applications, Vol. 4, 30.05.2012, p. 71-77.

    Research output: Contribution to journalArticle

    TY - JOUR

    T1 - Learning Video Manifolds for Content Analysis of Crowded Scenes

    AU - Thida, Myo

    AU - How-Lung, Eng

    AU - Monekosso, Dorothy

    AU - Remagnino, Paolo

    PY - 2012/5/30

    Y1 - 2012/5/30

    N2 - In this paper, we propose a new approach for recognizing group events and abnormality detection in a crowded scene. A manifold learning algorithm with temporal-constraints is proposed to embed a video of a crowded scene in a low-dimensional space. Our low dimensional representation of a video preserves the spatial temporal property of a video as well as the characteristic of the video. Recognizing video events and abnormality detection in a crowded scene is achieved by studying the video trajectory in the manifold space. We evaluate our proposed method on the state-of-the-art public data-sets containing different crowd events. Qualitative and quantitative results show the promising performance of the proposed method.

    AB - In this paper, we propose a new approach for recognizing group events and abnormality detection in a crowded scene. A manifold learning algorithm with temporal-constraints is proposed to embed a video of a crowded scene in a low-dimensional space. Our low dimensional representation of a video preserves the spatial temporal property of a video as well as the characteristic of the video. Recognizing video events and abnormality detection in a crowded scene is achieved by studying the video trajectory in the manifold space. We evaluate our proposed method on the state-of-the-art public data-sets containing different crowd events. Qualitative and quantitative results show the promising performance of the proposed method.

    U2 - 10.2197/ipsjtcva.4.71

    DO - 10.2197/ipsjtcva.4.71

    M3 - Article

    VL - 4

    SP - 71

    EP - 77

    JO - IPSJ Transactions on Computer Vision and Applications

    T2 - IPSJ Transactions on Computer Vision and Applications

    JF - IPSJ Transactions on Computer Vision and Applications

    SN - 1882-6695

    ER -