Detecting Photoshopped Faces by Scripting Photoshop
This addresses the issue of malicious photo manipulation for security and media verification, representing a novel method for a known bottleneck.
The researchers tackled the problem of detecting Photoshop-manipulated faces by training a model on automatically generated fake images, achieving performance that surpasses human ability in recognizing edits and even enabling reconstruction of original images in some cases.
Most malicious photo manipulations are created using standard image editing tools, such as Adobe Photoshop. We present a method for detecting one very popular Photoshop manipulation -- image warping applied to human faces -- using a model trained entirely using fake images that were automatically generated by scripting Photoshop itself. We show that our model outperforms humans at the task of recognizing manipulated images, can predict the specific location of edits, and in some cases can be used to "undo" a manipulation to reconstruct the original, unedited image. We demonstrate that the system can be successfully applied to real, artist-created image manipulations.