Startup unveils video-narrator generator from a single photo

The Israeli AI startup D-ID has launched the Creative Reality Studio platform to create a video with a voiceover from a single image.

Users need to upload a photo of the speaker or choose one of the available options. They can then insert the text that needs to be spoken, or upload a ready-made audio track with speech.

The developers said the platform supports 119 languages. Users can customize the voice and set its mood.

Based on the given parameters, the Creative Reality Studio algorithms generate a video in which the synthesized narrator reads the specified text.

According to the developers, the video generation time is half the length of the clip. During testing, however, journalists noted that creating a 60-second speech took several minutes.

The company believes their service will be in demand in sectors such as corporate training and education, internal and external communications, marketing and sales.

“We use our AI to create hosts and mentors who reproduce people and make content more engaging and effective,” said D-ID CEO Gil Perry.

To avoid creating deepfakes, the developers have imposed a number of restrictions. Users cannot upload profanity or racist statements, as well as photos of famous people. The platform’s rules prohibit creating videos with political content.

If terms of use are violated, the company may suspend the offender’s account and remove their video from the library.

A 14-day trial is available for new accounts. After two weeks, users can subscribe for $49 per month and generate Full HD videos with a total duration of 15 minutes.

In March 2022, D-ID and MyHeritage taught photos to talk.

In October 2021, the Israeli startup developed the Speaking Portraits tool, which animates a person in a portrait photo.

Subscribe to ForkLog AI news on Telegram: ForkLog AI — all the news from the world of AI!