The photos here are AI-generated clips from the Vidu website. This tool can create videos from text or image prompts.
Evelyn Chen | CNBC
BEIJING – Beijing-based Shengshu Technology announced Wednesday that its artificial intelligence-powered text-to-video conversion tool Vidu can now combine images to generate videos.
Vidu already allows users around the world to create 8-second clips based on written prompts. OpenAI, the developer of ChatGPT, revealed in February that its AI model Sora can generate one-minute videos from text, but has not yet made it publicly available.
Vidu’s new AI feature can combine three photos, such as a shirt, a person, and a moped, to create a video of the person wearing the shirt driving the moped in the field, Shengshu said.
Other platforms claim to be able to convert text and images into videos using AI, but the quality of the output varies. The breakthrough, Shengshu claims, is the ability to take three unique images and integrate them into an AI-generated video with visual consistency.
“Very early on we identified [visual consistency] We saw this as a problem and wanted to solve it well,” Fan Bao, Shengshu’s chief technology officer, said in Mandarin, as translated by CNBC.
Vidu launched in April and went viral on TikTok for its ability to turn two profile photos into lifelike videos of people hugging.
The AI ​​video generator has already earned revenue from advertisers, animators and other companies, Shengshu co-founder and CEO Jiayu Tang said in Mandarin, as translated by CNBC. Monthly fees per customer will range from 100,000 yuan to 1 million yuan ($13,871 to $138,711), he said.
Tan said that to address copyright issues, companies may enter into contracts with artists to allow AI to imitate their painting style in advertisements. He said he has not seen any significant litigation involving consumer use of images.
Tang added that Vidu does not allow ordinary people to generate content using images of celebrities or “sensitive” individuals. He said the AI ​​tool also prohibits nudity and violent images. When it comes to personal photos, Tan said Vidu destroys the data in accordance with the global benchmark General Data Protection Regulation.
According to PitchBook, Shengshu was founded last year with support from Baidu Ventures, Alibaba-affiliated Ant Group, Chinese startup Zhipu AI, Qiming Venture Partners, the city of Beijing and others.
Tan said Vidu’s AI runs on cloud servers rented in China and overseas.
