Dr Simmi Dutta, Aditya Sharma, Ruban Nazir Shah, Harsharan Singh Raina
Keywords:
Deepfakes, FaceForensics, Resnet, XceptionNet
Abstract
The free access to large-scale public databases, alongside the fast progress of deep learning techniques, especially Generative Adversarial Networks, have led to the generation of very realistic fake contents which can be threatening and have various implications in today's world where enforcing fake news is pretty simple. This survey provides a radical review of techniques for manipulating face images including DeepFake methods, and methods to detect such manipulations. These techniques are, face synthesis, face identity swap, facial attributes manipulation, and countenance manipulation. For each manipulation type, we offer details regarding manipulation techniques, existing public databases, and key benchmarks for technology evaluation of faux detection methods, including a summary of results from those evaluations.
