Two-Stream Neural Networks for Tampered Face Detection
This work addresses face tampering detection, which is important for security and media integrity, but it appears incremental as it builds on existing methods with a new dataset.
The authors tackled the problem of detecting tampered faces by proposing a two-stream neural network, achieving effective results on a newly created dataset of 2010 tampered images.
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 swapping 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 effectiveness of our method.