Detecting Photoshopped Faces by Scripting Photoshop

Abstract

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.

References in corpus (7)

• Adam: A Method for Stochastic Optimization
• Deep Residual Learning for Image Recognition
• Pixel Recurrent Neural Networks
• DeMoN: Depth and Motion Network for Learning Monocular Stereo
• FaceForensics: A Large-scale Video Dataset for Forgery Detection in Human Faces
• Aligned and Non-Aligned Double JPEG Detection Using Convolutional Neural Networks
• Video-to-Video Synthesis

Cited by in corpus (9)

• Media Forensics and DeepFakes: an overview
• Exposing Fake Images with Forensic Similarity Graphs
• FakeSpotter: A Simple yet Robust Baseline for Spotting AI-Synthesized Fake Faces
• DeepFake Detection Based on the Discrepancy Between the Face and its Context
• GANprintR: Improved Fakes and Evaluation of the State of the Art in Face Manipulation Detection
• MMSys'21 Grand Challenge on Detecting Cheapfakes
• Zooming into Face Forensics: A Pixel-level Analysis
• The 2021 Image Similarity Dataset and Challenge
• Identifying Human Edited Images using a CNN