Google researchers have developed their latest AI, VLOGGER, which creates realistic
talking, moving and gesturing people from photographs. This innovation allows the AI to
create a short video of a character speaking with appropriate facial expressions and
gestures by adding an audio recording alongside a photo.
The applications of this new technology from Google are far-reaching. For example, it
could be used to create three-dimensional, realistic avatars for games and virtual reality.
Also, some artists could use the technology to digitize themselves and sell their
three-dimensional models to film studios. Google's technology could also be used to
create realistic chatbots that can interact with humans.
The potential uses for VLOGGER are also quite diverse. For example, the technology could
be very useful for presentations, training programs and communication on
low-bandwidth networks. It can also provide a solution as a virtual narrator or
human-computer text communication interface.
However, possible uses of this technology include the creation of digital deep forgeries.
This has raised concerns among legislators and security experts around the world.
VLOGGER also has existing limitations. Currently, the videos created are quite short and
the background is often static. Also, the characters' ability to move is limited, and their
gestures and speech, while realistic, can give away the work of the AI. However, despite
these limitations, it should be noted that there has been considerable progress in image
animation technology.