Perhaps in one the world’s most advanced tech research, a tech group has come up with an amazing work. This could be the way future of videos may shape into.
Researchers from the University of Washington have developed a system to graft audio-synced mouths onto existing videos. Scary, ehh!
The video shows how an audio from old videos (Former President Obama as a test case here) is grafted into a separate video with perfect mouth sync.
These researchers have used 14 hours of Obama’s weekly address videos to form a program of new video. Once the program was trained it was able to take audio clip from the former president, create realistic mouth shapes and synced the audio making it a realistic look alike video.
The research team will present a paper (PDF) on their project at ACM Transactions on Graphics.