Learning scale-variant and scale-invariant features for deep image classification

Nanne van Noord*, Eric Postma

*Corresponding author for this work

    Research output: Contribution to journalArticleScientificpeer-review

    30 Downloads (Pure)

    Abstract

    Convolutional Neural Networks (CNNs) require large image corpora to be trained on classification tasks. The variation in image resolutions, sizes of objects and patterns depicted, and image scales, hampers CNN training and performance, because the task-relevant information varies over spatial scales. Previous work attempting to deal with such scale variations focused on encouraging scale-invariant CNN representations. However, scale-invariant representations are incomplete representations of images, because images contain scale-variant information as well. This paper addresses the combined development of scale-invariant and scale-variant representations. We propose a multi-scale CNN method to encourage the recognition of both types of features and evaluate it on a challenging image classification task involving task-relevant characteristics at multiple scales. The results show that our multi-scale CNN outperforms single-scale CNN. This leads to the conclusion that encouraging the combined development of a scale-invariant and scale-variant representation in CNNs is beneficial to image recognition performance. (C) 2016 The Authors. Published by Elsevier Ltd.

    Original languageEnglish
    Pages (from-to)583-592
    Number of pages10
    JournalPattern Recognition
    Volume61
    DOIs
    Publication statusPublished - Jan 2017

    Keywords

    • Convolutional Neural Networks
    • Multi-scale
    • Artist Attribution
    • Scale-variant Features
    • VAN GOGH

    Cite this

    @article{32762f900578471596879f61d5a3cb99,
    title = "Learning scale-variant and scale-invariant features for deep image classification",
    abstract = "Convolutional Neural Networks (CNNs) require large image corpora to be trained on classification tasks. The variation in image resolutions, sizes of objects and patterns depicted, and image scales, hampers CNN training and performance, because the task-relevant information varies over spatial scales. Previous work attempting to deal with such scale variations focused on encouraging scale-invariant CNN representations. However, scale-invariant representations are incomplete representations of images, because images contain scale-variant information as well. This paper addresses the combined development of scale-invariant and scale-variant representations. We propose a multi-scale CNN method to encourage the recognition of both types of features and evaluate it on a challenging image classification task involving task-relevant characteristics at multiple scales. The results show that our multi-scale CNN outperforms single-scale CNN. This leads to the conclusion that encouraging the combined development of a scale-invariant and scale-variant representation in CNNs is beneficial to image recognition performance. (C) 2016 The Authors. Published by Elsevier Ltd.",
    keywords = "Convolutional Neural Networks, Multi-scale, Artist Attribution, Scale-variant Features, VAN GOGH",
    author = "{van Noord}, Nanne and Eric Postma",
    year = "2017",
    month = "1",
    doi = "10.1016/j.patcog.2016.06.005",
    language = "English",
    volume = "61",
    pages = "583--592",
    journal = "Pattern Recognition Letters",
    issn = "0167-8655",
    publisher = "ELSEVIER SCI LTD",

    }

    Learning scale-variant and scale-invariant features for deep image classification. / van Noord, Nanne; Postma, Eric.

    In: Pattern Recognition, Vol. 61, 01.2017, p. 583-592.

    Research output: Contribution to journalArticleScientificpeer-review

    TY - JOUR

    T1 - Learning scale-variant and scale-invariant features for deep image classification

    AU - van Noord, Nanne

    AU - Postma, Eric

    PY - 2017/1

    Y1 - 2017/1

    N2 - Convolutional Neural Networks (CNNs) require large image corpora to be trained on classification tasks. The variation in image resolutions, sizes of objects and patterns depicted, and image scales, hampers CNN training and performance, because the task-relevant information varies over spatial scales. Previous work attempting to deal with such scale variations focused on encouraging scale-invariant CNN representations. However, scale-invariant representations are incomplete representations of images, because images contain scale-variant information as well. This paper addresses the combined development of scale-invariant and scale-variant representations. We propose a multi-scale CNN method to encourage the recognition of both types of features and evaluate it on a challenging image classification task involving task-relevant characteristics at multiple scales. The results show that our multi-scale CNN outperforms single-scale CNN. This leads to the conclusion that encouraging the combined development of a scale-invariant and scale-variant representation in CNNs is beneficial to image recognition performance. (C) 2016 The Authors. Published by Elsevier Ltd.

    AB - Convolutional Neural Networks (CNNs) require large image corpora to be trained on classification tasks. The variation in image resolutions, sizes of objects and patterns depicted, and image scales, hampers CNN training and performance, because the task-relevant information varies over spatial scales. Previous work attempting to deal with such scale variations focused on encouraging scale-invariant CNN representations. However, scale-invariant representations are incomplete representations of images, because images contain scale-variant information as well. This paper addresses the combined development of scale-invariant and scale-variant representations. We propose a multi-scale CNN method to encourage the recognition of both types of features and evaluate it on a challenging image classification task involving task-relevant characteristics at multiple scales. The results show that our multi-scale CNN outperforms single-scale CNN. This leads to the conclusion that encouraging the combined development of a scale-invariant and scale-variant representation in CNNs is beneficial to image recognition performance. (C) 2016 The Authors. Published by Elsevier Ltd.

    KW - Convolutional Neural Networks

    KW - Multi-scale

    KW - Artist Attribution

    KW - Scale-variant Features

    KW - VAN GOGH

    U2 - 10.1016/j.patcog.2016.06.005

    DO - 10.1016/j.patcog.2016.06.005

    M3 - Article

    VL - 61

    SP - 583

    EP - 592

    JO - Pattern Recognition Letters

    JF - Pattern Recognition Letters

    SN - 0167-8655

    ER -