Startup Synthesia unveiled expressive AI-powered avatars

By: Bohdan Kaminskyi | 25.04.2024, 17:58

Synthesia

Nvidia-backed UK-based Synthesia has announced a new artificial intelligence-based expressive avatar technology. The digital characters are capable of conveying human emotions using text input.

Here's What We Know

According to the developers, avatars blur the boundaries between the virtual world and reality. The technology allows you to create professional videos without the involvement of actors, film crews and lengthy editing. To train the system in the Synthesia studio in London, actors read scripts in front of a green screen.

In the company's demonstration, an avatar reacted with appropriate emotions to the text "I'm happy," "I'm sad," and "I'm upset." Over 55,000 companies, including half of the Fortune 100, already use Synthesia avatars for corporate presentations and training videos.

The startup is taking action against the use of technology to create fakes and misinformation. Customers undergo strict verification and the videos are moderated.

Synthesia has also joined the Content Verification Coalition, which promotes digital "watermarks" to identify AI content.

Source: CNBC