This video has the goal of introducing deepfakes, showing state-of-the-art implementations in the area and examples of what this technology is capable of. It was part of a project developed for my EIT Digital Master's Degree.
📜 Abstract
Deep Learning algorithms are being used in several fields nowadays, such as Self-driving cars, Healthcare or Voice-Activated Assistants. Programs developed with these algorithms are getting easier for not experienced users to use and their features are constantly growing. Despite this being positive for contributing to previous sectors, deep learning-powered applications that could cause threats in politics, privacy and security are being developed too. An example of this is the so-called "deep fakes". This video has the goal of introducing this type of algorithms, showing state-of-the-art implementations in the area and examples of what this technology is capable of.
🔥 State of the art implementations
✦ First Order Motion Model for Image Animation: [ Ссылка ]
✦ Motion-supervised Co-Part Segmentation: [ Ссылка ]
✦ FaceSwap: [ Ссылка ]
✦ DeepFaceLab: [ Ссылка ]
📚 Full research: [ Ссылка ]
