Model Principle
Stable Video Diffusion is based on the stable diffusion technique and achieves the ability to generate high-quality videos by exploring the latent space and morphing between text cues.
Model Features
Customizable frame rate, ranging from 3 to 30 frames per second
Suitable for various video applications, such as multi-view synthesis of a single image, etc.
Open source video generation model denmark mobile database Stable Video Diffusion
Stability AI recently released an open-source video generation model, Stable Video Diffusion. The model is based on the company's existing Stable Diffusion text-to-image model and can convert existing images into short videos by animating them.
Model Principle
Stable Video Diffusion generates videos by exploring the latent space and morphing between textual cues based on the stable diffusion technique.
Model features:
Can generate 14 or 25 fps video
Customizable frame rate, ranging from 3 to 30 frames per second
Suitable for various video applications, such as multi-view synthesis of a single image, etc.
Application of Stable Video Diffusion.