And you can do all kinds of movements, dancing and something are all small cases…
The most amazing thing is that their model can understand the principles of physical motion in the real world, so the video is very realistic.
Not only that, it can also directly generate videos from text, mix various characters and replace actions…
Its core technology is based on the JST-1 model. JST-1 is the first video-3D basic model with practical physical understanding capabilities. It can allow any character to move in a specified manner according to user needs.
Core functions:
1. Controllable video generation: Users can describe the details of the specified character’s actions and scenes in text, and Viggle will generate videos based on these instructions.
2. Physics-based animation: A distinctive feature of the JST-1 model is its understanding of the principles of physics, which means that the generated video not only looks real, but also the characters ‘movements and interactions conform to actual physical laws. This improves the quality and realism of the video.
3. 3D character and scene creation: Viggle is not limited to traditional 2D video production, it can also create 3D characters and scenes.
Official website:http://viggle.ai
You can go to their discord experience for free:https://discord.gg/5kk5SKwTWd
@ViggleAI
Video: