Company’s startup is developing a fully programmable virtual narrator For this purpose, they use video and audio recordings of Reuters news releases and artificial intelligence, specifically technology in the Deepfake style. As a result, they have a virtual presenter who looks and sounds like a person making a sports review. Real-time football match data reading allows news to be streamed immediately, without scripting or editing.
The basic idea is that you can broadcast dozens of different “live” videos, simultaneously showing the same “person” who announces real-time updates for different sporting events.
In addition to news reporting, it is easy to imagine airports filled with monitors showing the same person who announces updates for flights of different airlines, or any number of the same videos but adapted to specific geographical areas.
As well as streamlining voice and audience loyalty, this technology can be used to create emergency video reports in places where language barriers may slow the spread of information, both orally and in writing.
Reuters notes that this is just a prototype. It is conceived as a proof of the potential of future real-time news and information services based on artificial intelligence systems. Earlier in China, the news was already run by a virtual narrator.
You can create your own small video based on artificial intelligence on the Synthesia startup website.