Learning-by-Synthesis for Appearance-Based 3D Gaze Estimation

Yusuke Sugano, Yasuyuki Matsushita, Yoichi Sato

CVPR 2014

doi:10.1109/CVPR.2014.235 /cvpr/2014/sugano2014cvpr-learningbysynthesis/

Abstract

Inferring human gaze from low-resolution eye images is still a challenging task despite its practical importance in many application scenarios. This paper presents a learning-by-synthesis approach to accurate image-based gaze estimation that is person- and head pose-independent. Unlike existing appearance-based methods that assume person-specific training data, we use a large amount of cross-subject training data to train a 3D gaze estimator. We collect the largest and fully calibrated multi-view gaze dataset and perform a 3D reconstruction in order to generate dense training data of eye images. By using the synthesized dataset to learn a random regression forest, we show that our method outperforms existing methods that use low-resolution eye images.

PDF CVPR Semantic Scholar

Cite

Text

Sugano et al. "Learning-by-Synthesis for Appearance-Based 3D Gaze Estimation." Conference on Computer Vision and Pattern Recognition, 2014. doi:10.1109/CVPR.2014.235

Markdown

[Sugano et al. "Learning-by-Synthesis for Appearance-Based 3D Gaze Estimation." Conference on Computer Vision and Pattern Recognition, 2014.](https://mlanthology.org/cvpr/2014/sugano2014cvpr-learningbysynthesis/) doi:10.1109/CVPR.2014.235

BibTeX

@inproceedings{sugano2014cvpr-learningbysynthesis,
  title     = {{Learning-by-Synthesis for Appearance-Based 3D Gaze Estimation}},
  author    = {Sugano, Yusuke and Matsushita, Yasuyuki and Sato, Yoichi},
  booktitle = {Conference on Computer Vision and Pattern Recognition},
  year      = {2014},
  doi       = {10.1109/CVPR.2014.235},
  url       = {https://mlanthology.org/cvpr/2014/sugano2014cvpr-learningbysynthesis/}
}