Towards automatic detection of child pornography

Napa Sae-Bae, Xiaoxi Sun, Husrev T. Sencar, Nasir D. Memon

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

    Abstract

    This paper presents a child pornographic image detection system that identifies human skin tones in digital images, extracts features to detect explicit images and performs facial image based age classification. The novelty of the technique relies on the use of a robust and very fast skin color filter and a new set of facial features for improved identification of child faces. Tests on a dataset containing explicit images taken under different illuminations and reflecting a diversity of human skin tones, show that explicit images can be differentiated from benign images with around 90% accuracy. Similarly, tests performed on adult and child facial images yielded an accuracy of 80% in detecting child faces. Test conducted on 105 images involving semi-naked children (with no sexual context) revealed that the system has true positive rates of 83% in detecting explicit-like images and 96.5% in detecting child faces.

    Original languageEnglish (US)
    Title of host publication2014 IEEE International Conference on Image Processing, ICIP 2014
    PublisherInstitute of Electrical and Electronics Engineers Inc.
    Pages5332-5336
    Number of pages5
    ISBN (Print)9781479957514
    DOIs
    StatePublished - Jan 28 2014

    Fingerprint

    Skin
    Lighting
    Color

    ASJC Scopus subject areas

    • Computer Vision and Pattern Recognition

    Cite this

    Sae-Bae, N., Sun, X., Sencar, H. T., & Memon, N. D. (2014). Towards automatic detection of child pornography. In 2014 IEEE International Conference on Image Processing, ICIP 2014 (pp. 5332-5336). [7026079] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICIP.2014.7026079

    Towards automatic detection of child pornography. / Sae-Bae, Napa; Sun, Xiaoxi; Sencar, Husrev T.; Memon, Nasir D.

    2014 IEEE International Conference on Image Processing, ICIP 2014. Institute of Electrical and Electronics Engineers Inc., 2014. p. 5332-5336 7026079.

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

    Sae-Bae, N, Sun, X, Sencar, HT & Memon, ND 2014, Towards automatic detection of child pornography. in 2014 IEEE International Conference on Image Processing, ICIP 2014., 7026079, Institute of Electrical and Electronics Engineers Inc., pp. 5332-5336. https://doi.org/10.1109/ICIP.2014.7026079
    Sae-Bae N, Sun X, Sencar HT, Memon ND. Towards automatic detection of child pornography. In 2014 IEEE International Conference on Image Processing, ICIP 2014. Institute of Electrical and Electronics Engineers Inc. 2014. p. 5332-5336. 7026079 https://doi.org/10.1109/ICIP.2014.7026079
    Sae-Bae, Napa ; Sun, Xiaoxi ; Sencar, Husrev T. ; Memon, Nasir D. / Towards automatic detection of child pornography. 2014 IEEE International Conference on Image Processing, ICIP 2014. Institute of Electrical and Electronics Engineers Inc., 2014. pp. 5332-5336
    @inproceedings{dd7cee131239432fb3d05605c656c82b,
    title = "Towards automatic detection of child pornography",
    abstract = "This paper presents a child pornographic image detection system that identifies human skin tones in digital images, extracts features to detect explicit images and performs facial image based age classification. The novelty of the technique relies on the use of a robust and very fast skin color filter and a new set of facial features for improved identification of child faces. Tests on a dataset containing explicit images taken under different illuminations and reflecting a diversity of human skin tones, show that explicit images can be differentiated from benign images with around 90{\%} accuracy. Similarly, tests performed on adult and child facial images yielded an accuracy of 80{\%} in detecting child faces. Test conducted on 105 images involving semi-naked children (with no sexual context) revealed that the system has true positive rates of 83{\%} in detecting explicit-like images and 96.5{\%} in detecting child faces.",
    author = "Napa Sae-Bae and Xiaoxi Sun and Sencar, {Husrev T.} and Memon, {Nasir D.}",
    year = "2014",
    month = "1",
    day = "28",
    doi = "10.1109/ICIP.2014.7026079",
    language = "English (US)",
    isbn = "9781479957514",
    pages = "5332--5336",
    booktitle = "2014 IEEE International Conference on Image Processing, ICIP 2014",
    publisher = "Institute of Electrical and Electronics Engineers Inc.",

    }

    TY - GEN

    T1 - Towards automatic detection of child pornography

    AU - Sae-Bae, Napa

    AU - Sun, Xiaoxi

    AU - Sencar, Husrev T.

    AU - Memon, Nasir D.

    PY - 2014/1/28

    Y1 - 2014/1/28

    N2 - This paper presents a child pornographic image detection system that identifies human skin tones in digital images, extracts features to detect explicit images and performs facial image based age classification. The novelty of the technique relies on the use of a robust and very fast skin color filter and a new set of facial features for improved identification of child faces. Tests on a dataset containing explicit images taken under different illuminations and reflecting a diversity of human skin tones, show that explicit images can be differentiated from benign images with around 90% accuracy. Similarly, tests performed on adult and child facial images yielded an accuracy of 80% in detecting child faces. Test conducted on 105 images involving semi-naked children (with no sexual context) revealed that the system has true positive rates of 83% in detecting explicit-like images and 96.5% in detecting child faces.

    AB - This paper presents a child pornographic image detection system that identifies human skin tones in digital images, extracts features to detect explicit images and performs facial image based age classification. The novelty of the technique relies on the use of a robust and very fast skin color filter and a new set of facial features for improved identification of child faces. Tests on a dataset containing explicit images taken under different illuminations and reflecting a diversity of human skin tones, show that explicit images can be differentiated from benign images with around 90% accuracy. Similarly, tests performed on adult and child facial images yielded an accuracy of 80% in detecting child faces. Test conducted on 105 images involving semi-naked children (with no sexual context) revealed that the system has true positive rates of 83% in detecting explicit-like images and 96.5% in detecting child faces.

    UR - http://www.scopus.com/inward/record.url?scp=84949928560&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84949928560&partnerID=8YFLogxK

    U2 - 10.1109/ICIP.2014.7026079

    DO - 10.1109/ICIP.2014.7026079

    M3 - Conference contribution

    SN - 9781479957514

    SP - 5332

    EP - 5336

    BT - 2014 IEEE International Conference on Image Processing, ICIP 2014

    PB - Institute of Electrical and Electronics Engineers Inc.

    ER -