Deep Learning with Feature Reuse for JPEG Image Steganalysis

Jianhua Yang, Xiangui Kang, Edward Wong, Yun Qing Shi

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Abstract

    It is challenging to detect weak hidden information in a JPEG compressed image. In this paper, we propose a 32-layer convolutional neural networks (CNNs) with feature reuse by concatenating all features from previous layers. The proposed method can improve the flow of gradient and information, and the shared features and bottleneck layers in the proposed CNN model further reduce the number of parameters dramatically. The experimental results shown that the proposed method significantly reduce the detection error rate compared with the existing JPEG steganalysis methods, e.g. state-of-the-art XuNet method and the conventional SCA-GFR method. Compared with XuNet method and conventional method SCA-GFR in detecting J-UNIWARD at 0.1 bpnzAC (bit per non-zero AC DCT coefficient), the proposed method can reduce detection error rate by 4.33% and 6.55% respectively.

    Original languageEnglish (US)
    Title of host publication2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018 - Proceedings
    PublisherInstitute of Electrical and Electronics Engineers Inc.
    Pages533-538
    Number of pages6
    ISBN (Electronic)9789881476852
    DOIs
    StatePublished - Mar 4 2019
    Event10th Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018 - Honolulu, United States
    Duration: Nov 12 2018Nov 15 2018

    Publication series

    Name2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018 - Proceedings

    Conference

    Conference10th Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018
    CountryUnited States
    CityHonolulu
    Period11/12/1811/15/18

    Fingerprint

    Error detection
    Neural networks
    Deep learning

    ASJC Scopus subject areas

    • Information Systems

    Cite this

    Yang, J., Kang, X., Wong, E., & Shi, Y. Q. (2019). Deep Learning with Feature Reuse for JPEG Image Steganalysis. In 2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018 - Proceedings (pp. 533-538). [8659589] (2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018 - Proceedings). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.23919/APSIPA.2018.8659589

    Deep Learning with Feature Reuse for JPEG Image Steganalysis. / Yang, Jianhua; Kang, Xiangui; Wong, Edward; Shi, Yun Qing.

    2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2019. p. 533-538 8659589 (2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018 - Proceedings).

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Yang, J, Kang, X, Wong, E & Shi, YQ 2019, Deep Learning with Feature Reuse for JPEG Image Steganalysis. in 2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018 - Proceedings., 8659589, 2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018 - Proceedings, Institute of Electrical and Electronics Engineers Inc., pp. 533-538, 10th Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018, Honolulu, United States, 11/12/18. https://doi.org/10.23919/APSIPA.2018.8659589
    Yang J, Kang X, Wong E, Shi YQ. Deep Learning with Feature Reuse for JPEG Image Steganalysis. In 2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018 - Proceedings. Institute of Electrical and Electronics Engineers Inc. 2019. p. 533-538. 8659589. (2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018 - Proceedings). https://doi.org/10.23919/APSIPA.2018.8659589
    Yang, Jianhua ; Kang, Xiangui ; Wong, Edward ; Shi, Yun Qing. / Deep Learning with Feature Reuse for JPEG Image Steganalysis. 2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2019. pp. 533-538 (2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018 - Proceedings).
    @inproceedings{70ef60d1300c435b9fa67088ed8e8f3b,
    title = "Deep Learning with Feature Reuse for JPEG Image Steganalysis",
    abstract = "It is challenging to detect weak hidden information in a JPEG compressed image. In this paper, we propose a 32-layer convolutional neural networks (CNNs) with feature reuse by concatenating all features from previous layers. The proposed method can improve the flow of gradient and information, and the shared features and bottleneck layers in the proposed CNN model further reduce the number of parameters dramatically. The experimental results shown that the proposed method significantly reduce the detection error rate compared with the existing JPEG steganalysis methods, e.g. state-of-the-art XuNet method and the conventional SCA-GFR method. Compared with XuNet method and conventional method SCA-GFR in detecting J-UNIWARD at 0.1 bpnzAC (bit per non-zero AC DCT coefficient), the proposed method can reduce detection error rate by 4.33{\%} and 6.55{\%} respectively.",
    author = "Jianhua Yang and Xiangui Kang and Edward Wong and Shi, {Yun Qing}",
    year = "2019",
    month = "3",
    day = "4",
    doi = "10.23919/APSIPA.2018.8659589",
    language = "English (US)",
    series = "2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018 - Proceedings",
    publisher = "Institute of Electrical and Electronics Engineers Inc.",
    pages = "533--538",
    booktitle = "2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018 - Proceedings",

    }

    TY - GEN

    T1 - Deep Learning with Feature Reuse for JPEG Image Steganalysis

    AU - Yang, Jianhua

    AU - Kang, Xiangui

    AU - Wong, Edward

    AU - Shi, Yun Qing

    PY - 2019/3/4

    Y1 - 2019/3/4

    N2 - It is challenging to detect weak hidden information in a JPEG compressed image. In this paper, we propose a 32-layer convolutional neural networks (CNNs) with feature reuse by concatenating all features from previous layers. The proposed method can improve the flow of gradient and information, and the shared features and bottleneck layers in the proposed CNN model further reduce the number of parameters dramatically. The experimental results shown that the proposed method significantly reduce the detection error rate compared with the existing JPEG steganalysis methods, e.g. state-of-the-art XuNet method and the conventional SCA-GFR method. Compared with XuNet method and conventional method SCA-GFR in detecting J-UNIWARD at 0.1 bpnzAC (bit per non-zero AC DCT coefficient), the proposed method can reduce detection error rate by 4.33% and 6.55% respectively.

    AB - It is challenging to detect weak hidden information in a JPEG compressed image. In this paper, we propose a 32-layer convolutional neural networks (CNNs) with feature reuse by concatenating all features from previous layers. The proposed method can improve the flow of gradient and information, and the shared features and bottleneck layers in the proposed CNN model further reduce the number of parameters dramatically. The experimental results shown that the proposed method significantly reduce the detection error rate compared with the existing JPEG steganalysis methods, e.g. state-of-the-art XuNet method and the conventional SCA-GFR method. Compared with XuNet method and conventional method SCA-GFR in detecting J-UNIWARD at 0.1 bpnzAC (bit per non-zero AC DCT coefficient), the proposed method can reduce detection error rate by 4.33% and 6.55% respectively.

    UR - http://www.scopus.com/inward/record.url?scp=85063439055&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=85063439055&partnerID=8YFLogxK

    U2 - 10.23919/APSIPA.2018.8659589

    DO - 10.23919/APSIPA.2018.8659589

    M3 - Conference contribution

    AN - SCOPUS:85063439055

    T3 - 2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018 - Proceedings

    SP - 533

    EP - 538

    BT - 2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018 - Proceedings

    PB - Institute of Electrical and Electronics Engineers Inc.

    ER -