3D facial geometric features for constrained local model

Shiyang Cheng, Stefanos Zafeiriou, Ashish Asthana, Akshay Asthana, Maja Pantic

    Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

    13 Citations (Scopus)
    87 Downloads (Pure)

    Abstract

    We propose a 3D Constrained Local Model framework for deformable face alignment in depth image. Our framework exploits the intrinsic 3D geometric information in depth data by utilizing robust histogram-based 3D geometric features that are based on normal vectors. In addition, we demonstrate the fusion of intensity data and 3D features that further improves the facial landmark localization accuracy. The experiments are conducted on publicly available FRGC database. The results show that our 3D features based CLM completely outperforms the raw depth features based CLM in term of fitting accuracy and robustness, and the fusion of intensity and 3D depth feature further improves the performance. Another benefit is that the proposed 3D features in our framework do not require any pre-processing procedure on the data.
    Original languageUndefined
    Title of host publicationProceedings of IEEE International Conference on Image Processing (ICIP 2014)
    Place of PublicationUSA
    PublisherIEEE Computer Society
    Pages1425-1429
    Number of pages5
    ISBN (Print)978-1-4799-5751-4
    DOIs
    Publication statusPublished - Oct 2014
    EventIEEE International Conference on Image Processing 2014 - Paris, France, Paris, France
    Duration: 27 Oct 201430 Oct 2014
    https://icip2014.wp.imt.fr/

    Publication series

    Name
    PublisherIEEE Computer Society

    Conference

    ConferenceIEEE International Conference on Image Processing 2014
    Abbreviated titleICIP 2014
    CountryFrance
    CityParis
    Period27/10/1430/10/14
    Internet address

    Keywords

    • EWI-25826
    • HMI-HF: Human Factors
    • deformable face alignment
    • IR-95232
    • Constrained Local Model
    • 3D facial geometry
    • METIS-309951
    • histogram-based 3D feature

    Cite this

    Cheng, S., Zafeiriou, S., Asthana, A., Asthana, A., & Pantic, M. (2014). 3D facial geometric features for constrained local model. In Proceedings of IEEE International Conference on Image Processing (ICIP 2014) (pp. 1425-1429). USA: IEEE Computer Society. https://doi.org/10.1109/ICIP.2014.7025285
    Cheng, Shiyang ; Zafeiriou, Stefanos ; Asthana, Ashish ; Asthana, Akshay ; Pantic, Maja. / 3D facial geometric features for constrained local model. Proceedings of IEEE International Conference on Image Processing (ICIP 2014). USA : IEEE Computer Society, 2014. pp. 1425-1429
    @inproceedings{a53174623b3e4087be1c39ed9d4665dc,
    title = "3D facial geometric features for constrained local model",
    abstract = "We propose a 3D Constrained Local Model framework for deformable face alignment in depth image. Our framework exploits the intrinsic 3D geometric information in depth data by utilizing robust histogram-based 3D geometric features that are based on normal vectors. In addition, we demonstrate the fusion of intensity data and 3D features that further improves the facial landmark localization accuracy. The experiments are conducted on publicly available FRGC database. The results show that our 3D features based CLM completely outperforms the raw depth features based CLM in term of fitting accuracy and robustness, and the fusion of intensity and 3D depth feature further improves the performance. Another benefit is that the proposed 3D features in our framework do not require any pre-processing procedure on the data.",
    keywords = "EWI-25826, HMI-HF: Human Factors, deformable face alignment, IR-95232, Constrained Local Model, 3D facial geometry, METIS-309951, histogram-based 3D feature",
    author = "Shiyang Cheng and Stefanos Zafeiriou and Ashish Asthana and Akshay Asthana and Maja Pantic",
    note = "eemcs-eprint-25826",
    year = "2014",
    month = "10",
    doi = "10.1109/ICIP.2014.7025285",
    language = "Undefined",
    isbn = "978-1-4799-5751-4",
    publisher = "IEEE Computer Society",
    pages = "1425--1429",
    booktitle = "Proceedings of IEEE International Conference on Image Processing (ICIP 2014)",
    address = "United States",

    }

    Cheng, S, Zafeiriou, S, Asthana, A, Asthana, A & Pantic, M 2014, 3D facial geometric features for constrained local model. in Proceedings of IEEE International Conference on Image Processing (ICIP 2014). IEEE Computer Society, USA, pp. 1425-1429, IEEE International Conference on Image Processing 2014, Paris, France, 27/10/14. https://doi.org/10.1109/ICIP.2014.7025285

    3D facial geometric features for constrained local model. / Cheng, Shiyang; Zafeiriou, Stefanos; Asthana, Ashish; Asthana, Akshay; Pantic, Maja.

    Proceedings of IEEE International Conference on Image Processing (ICIP 2014). USA : IEEE Computer Society, 2014. p. 1425-1429.

    Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

    TY - GEN

    T1 - 3D facial geometric features for constrained local model

    AU - Cheng, Shiyang

    AU - Zafeiriou, Stefanos

    AU - Asthana, Ashish

    AU - Asthana, Akshay

    AU - Pantic, Maja

    N1 - eemcs-eprint-25826

    PY - 2014/10

    Y1 - 2014/10

    N2 - We propose a 3D Constrained Local Model framework for deformable face alignment in depth image. Our framework exploits the intrinsic 3D geometric information in depth data by utilizing robust histogram-based 3D geometric features that are based on normal vectors. In addition, we demonstrate the fusion of intensity data and 3D features that further improves the facial landmark localization accuracy. The experiments are conducted on publicly available FRGC database. The results show that our 3D features based CLM completely outperforms the raw depth features based CLM in term of fitting accuracy and robustness, and the fusion of intensity and 3D depth feature further improves the performance. Another benefit is that the proposed 3D features in our framework do not require any pre-processing procedure on the data.

    AB - We propose a 3D Constrained Local Model framework for deformable face alignment in depth image. Our framework exploits the intrinsic 3D geometric information in depth data by utilizing robust histogram-based 3D geometric features that are based on normal vectors. In addition, we demonstrate the fusion of intensity data and 3D features that further improves the facial landmark localization accuracy. The experiments are conducted on publicly available FRGC database. The results show that our 3D features based CLM completely outperforms the raw depth features based CLM in term of fitting accuracy and robustness, and the fusion of intensity and 3D depth feature further improves the performance. Another benefit is that the proposed 3D features in our framework do not require any pre-processing procedure on the data.

    KW - EWI-25826

    KW - HMI-HF: Human Factors

    KW - deformable face alignment

    KW - IR-95232

    KW - Constrained Local Model

    KW - 3D facial geometry

    KW - METIS-309951

    KW - histogram-based 3D feature

    U2 - 10.1109/ICIP.2014.7025285

    DO - 10.1109/ICIP.2014.7025285

    M3 - Conference contribution

    SN - 978-1-4799-5751-4

    SP - 1425

    EP - 1429

    BT - Proceedings of IEEE International Conference on Image Processing (ICIP 2014)

    PB - IEEE Computer Society

    CY - USA

    ER -

    Cheng S, Zafeiriou S, Asthana A, Asthana A, Pantic M. 3D facial geometric features for constrained local model. In Proceedings of IEEE International Conference on Image Processing (ICIP 2014). USA: IEEE Computer Society. 2014. p. 1425-1429 https://doi.org/10.1109/ICIP.2014.7025285