Two-Stream Neural Networks for Tampered Face Detection

Abstract

We propose a two-stream network for face tampering detection. We train GoogLeNet to detect tampering artifacts in a face classification stream, and train a patch based triplet network to leverage features capturing local noise residuals and camera characteristics as a second stream. In addition, we use two different online face swaping applications to create a new dataset that consists of 2010 tampered images, each of which contains a tampered face. We evaluate the proposed two-stream network on our newly collected dataset. Experimental results demonstrate the effectness of our method.

Cite

Text

Zhou et al. "Two-Stream Neural Networks for Tampered Face Detection." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2017. doi:10.1109/CVPRW.2017.229

Markdown

[Zhou et al. "Two-Stream Neural Networks for Tampered Face Detection." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2017.](https://mlanthology.org/cvprw/2017/zhou2017cvprw-twostream/) doi:10.1109/CVPRW.2017.229

BibTeX

@inproceedings{zhou2017cvprw-twostream,
  title     = {{Two-Stream Neural Networks for Tampered Face Detection}},
  author    = {Zhou, Peng and Han, Xintong and Morariu, Vlad I. and Davis, Larry S.},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
  year      = {2017},
  pages     = {1831-1839},
  doi       = {10.1109/CVPRW.2017.229},
  url       = {https://mlanthology.org/cvprw/2017/zhou2017cvprw-twostream/}
}