Large Tech’s race to leapfrog the most recent AI fashions continues with the launch of ByteDance’s next-gen video generator. In a blog post, ByteDance – the China-based firm behind TikTok – says Seedance 2.0 helps prompts that mix textual content, pictures, video, and audio.
The corporate claims it “delivers a considerable leap in era high quality,” providing enhancements in producing complicated scenes with a number of topics and its skill to comply with directions. Customers can refine their textual content prompts by feeding Seedance 2.0 as much as 9 pictures, three video clips, and three audio clips.
The mannequin can generate as much as 15-second clips with audio, whereas taking digital camera motion, visible results, and movement under consideration. It may possibly additionally reference text-based storyboards, in keeping with ByteDance.
In a single instance shared by ByteDance, which reveals two determine skaters performing a routine collectively, the corporate says Seedance 2.0 can “reliably carry out a sequence of high-difficulty actions — together with synchronized takeoffs, mid-air spins and exact ice landings — whereas strictly following real-world bodily legal guidelines.”
Customers on social media have already began exhibiting off what the brand new software can do, with one individual posting an AI-generated video with the likenesses of Brad Pitt and Tom Cruise in a cinematic combat sequence. Deadpool author Rhett Reese reposted the video with the remark, “I hate to say it. It’s possible over for us.”
For now, Seedance 2.0 is barely obtainable by means of ByteDance’s Dreamina AI platform and thru its AI assistant, Doubao. It’s unclear whether or not it would make its strategy to TikTok — particularly now that the app is underneath new possession within the US.
