Iterative deep learning for image set based face and object recognition

    Research output: Contribution to journalArticle

    39 Citations (Scopus)

    Abstract

    We present a novel technique for image set based face/object recognition, where each gallery and query example contains a face/object image set captured from different viewpoints, background, facial expressions, resolution and illumination levels. While several image set classification approaches have been proposed in recent years, most of them represent each image set as a single linear subspace, mixture of linear subspaces or Lie group of Riemannian manifold. These techniques make prior assumptions in regards to the specific category of the geometric surface on which images of the set are believed to lie. This could result in a loss of discriminative information for classification. This paper alleviates these limitations by proposing an Iterative Deep Learning Model (IDLM) that automatically and hierarchically learns discriminative representations from raw face and object images. In the proposed approach, low level translationally invariant features are learnt by the Pooled Convolutional Layer (PCL). The latter is followed by Artificial Neural Networks (ANNs) applied iteratively in a hierarchical fashion to learn a discriminative non-linear feature representation of the input image sets. The proposed technique was extensively evaluated for the task of image set based face and object recognition on YouTube Celebrities, Honda/UCSD, CMU Mobo and ETH-80 (object) dataset, respectively. Experimental results and comparisons with state-of-the-art methods show that our technique achieves the best performance on all these datasets.
    Original languageEnglish
    Pages (from-to)866-874
    JournalNeurocomputing
    Volume174
    Issue numberPart B
    Early online date22 Oct 2015
    DOIs
    Publication statusPublished - 22 Jan 2016

    Fingerprint

    Object recognition
    Face recognition
    Learning
    Lie groups
    Facial Expression
    Lighting
    Neural networks
    Deep learning
    Facial Recognition
    Datasets

    Cite this

    @article{2167418affdd4e3d99e234c2f8dc60bf,
    title = "Iterative deep learning for image set based face and object recognition",
    abstract = "We present a novel technique for image set based face/object recognition, where each gallery and query example contains a face/object image set captured from different viewpoints, background, facial expressions, resolution and illumination levels. While several image set classification approaches have been proposed in recent years, most of them represent each image set as a single linear subspace, mixture of linear subspaces or Lie group of Riemannian manifold. These techniques make prior assumptions in regards to the specific category of the geometric surface on which images of the set are believed to lie. This could result in a loss of discriminative information for classification. This paper alleviates these limitations by proposing an Iterative Deep Learning Model (IDLM) that automatically and hierarchically learns discriminative representations from raw face and object images. In the proposed approach, low level translationally invariant features are learnt by the Pooled Convolutional Layer (PCL). The latter is followed by Artificial Neural Networks (ANNs) applied iteratively in a hierarchical fashion to learn a discriminative non-linear feature representation of the input image sets. The proposed technique was extensively evaluated for the task of image set based face and object recognition on YouTube Celebrities, Honda/UCSD, CMU Mobo and ETH-80 (object) dataset, respectively. Experimental results and comparisons with state-of-the-art methods show that our technique achieves the best performance on all these datasets.",
    author = "Shah, {Syed Afaq} and Mohammed Bennamoun and Farid Boussaid",
    year = "2016",
    month = "1",
    day = "22",
    doi = "10.1016/j.neucom.2015.10.004",
    language = "English",
    volume = "174",
    pages = "866--874",
    journal = "Neurocomputing",
    issn = "0925-2312",
    publisher = "Pergamon",
    number = "Part B",

    }

    Iterative deep learning for image set based face and object recognition. / Shah, Syed Afaq; Bennamoun, Mohammed; Boussaid, Farid.

    In: Neurocomputing, Vol. 174, No. Part B, 22.01.2016, p. 866-874.

    Research output: Contribution to journalArticle

    TY - JOUR

    T1 - Iterative deep learning for image set based face and object recognition

    AU - Shah, Syed Afaq

    AU - Bennamoun, Mohammed

    AU - Boussaid, Farid

    PY - 2016/1/22

    Y1 - 2016/1/22

    N2 - We present a novel technique for image set based face/object recognition, where each gallery and query example contains a face/object image set captured from different viewpoints, background, facial expressions, resolution and illumination levels. While several image set classification approaches have been proposed in recent years, most of them represent each image set as a single linear subspace, mixture of linear subspaces or Lie group of Riemannian manifold. These techniques make prior assumptions in regards to the specific category of the geometric surface on which images of the set are believed to lie. This could result in a loss of discriminative information for classification. This paper alleviates these limitations by proposing an Iterative Deep Learning Model (IDLM) that automatically and hierarchically learns discriminative representations from raw face and object images. In the proposed approach, low level translationally invariant features are learnt by the Pooled Convolutional Layer (PCL). The latter is followed by Artificial Neural Networks (ANNs) applied iteratively in a hierarchical fashion to learn a discriminative non-linear feature representation of the input image sets. The proposed technique was extensively evaluated for the task of image set based face and object recognition on YouTube Celebrities, Honda/UCSD, CMU Mobo and ETH-80 (object) dataset, respectively. Experimental results and comparisons with state-of-the-art methods show that our technique achieves the best performance on all these datasets.

    AB - We present a novel technique for image set based face/object recognition, where each gallery and query example contains a face/object image set captured from different viewpoints, background, facial expressions, resolution and illumination levels. While several image set classification approaches have been proposed in recent years, most of them represent each image set as a single linear subspace, mixture of linear subspaces or Lie group of Riemannian manifold. These techniques make prior assumptions in regards to the specific category of the geometric surface on which images of the set are believed to lie. This could result in a loss of discriminative information for classification. This paper alleviates these limitations by proposing an Iterative Deep Learning Model (IDLM) that automatically and hierarchically learns discriminative representations from raw face and object images. In the proposed approach, low level translationally invariant features are learnt by the Pooled Convolutional Layer (PCL). The latter is followed by Artificial Neural Networks (ANNs) applied iteratively in a hierarchical fashion to learn a discriminative non-linear feature representation of the input image sets. The proposed technique was extensively evaluated for the task of image set based face and object recognition on YouTube Celebrities, Honda/UCSD, CMU Mobo and ETH-80 (object) dataset, respectively. Experimental results and comparisons with state-of-the-art methods show that our technique achieves the best performance on all these datasets.

    U2 - 10.1016/j.neucom.2015.10.004

    DO - 10.1016/j.neucom.2015.10.004

    M3 - Article

    VL - 174

    SP - 866

    EP - 874

    JO - Neurocomputing

    JF - Neurocomputing

    SN - 0925-2312

    IS - Part B

    ER -