Challenges Related to Deepfakes
Learn about the challenges related to the deepfakes setup and a few notable implementations.
We'll cover the following...
Let’s discuss some of the common challenges associated with deepfake architectures, beginning with a brief discussion of the ethical issues associated with this technology.
Ethical issues
Even though generating fake content is not a new concept, the word “deepfake” came into the limelight in 2017 when a Reddit user posted fake videos with celebrity faces superimposed on them using deep learning. The quality of the content and the ease with which the user was able to generate them created a huge uproar on news channels across the globe. Soon, the Reddit user released an easy-to-setup application called FakeApp that enabled users to generate such content with very little knowledge of how deep learning works. This led to a number of fake videos and objectionable content. This, in turn, helped people gain traction on issues associated with identity theft, impersonation, fake news, and so on.
Soon, interest picked up within the academic community, which not only helped to improve the technology but also insisted on its ethical use. While malicious and objectionable content creators are using these techniques, a number of industry and research projects are underway to detect such fake content, such as Microsoft’s deepfake detection