ML Anthology

Facial Landmark Detection via Progressive Initialization

Shengtao Xiao, Shuicheng Yan, Ashraf A. Kassim
ICCVW 2015 pp. 986-993
doi:10.1109/ICCVW.2015.130 /iccvw/2015/xiao2015iccvw-facial/

Abstract

In this paper, we present a multi-stage regression-based approach for the 300 Videos in-the-Wild (300-VW) Challenge, which progressively initializes the shape from obvious landmarks with strong semantic meanings, e.g. eyes and mouth corners, to landmarks on face contour, eyebrows and nose bridge which have more challenging features. Compared with initialization based on mean shape and multiple random shapes, our proposed progressive initialization can very robustly handle challenging poses. It also guarantees an accurate landmark localization result and shows smooth tracking performance in real-time.

ICCVW Semantic Scholar

Cite

Text

Xiao et al. "Facial Landmark Detection via Progressive Initialization." IEEE/CVF International Conference on Computer Vision Workshops, 2015. doi:10.1109/ICCVW.2015.130

Markdown

[Xiao et al. "Facial Landmark Detection via Progressive Initialization." IEEE/CVF International Conference on Computer Vision Workshops, 2015.](https://mlanthology.org/iccvw/2015/xiao2015iccvw-facial/) doi:10.1109/ICCVW.2015.130

BibTeX

@inproceedings{xiao2015iccvw-facial,
  title     = {{Facial Landmark Detection via Progressive Initialization}},
  author    = {Xiao, Shengtao and Yan, Shuicheng and Kassim, Ashraf A.},
  booktitle = {IEEE/CVF International Conference on Computer Vision Workshops},
  year      = {2015},
  pages     = {986-993},
  doi       = {10.1109/ICCVW.2015.130},
  url       = {https://mlanthology.org/iccvw/2015/xiao2015iccvw-facial/}
}