AnimateDiff
An open-source local AI animation module that works with Stable Diffusion 1.5 models and supports camera motion controls and image or sketch guidance.
AnimateDiff generates animations with personalized Stable Diffusion models for artists and developers who want motion in their chosen visual style. Compatible image models don't need separate animation training.
- Personalized models: Animate Stable Diffusion 1.5 models customized with LoRA or DreamBooth, including ToonYou, Realistic Vision and majicMIX Realistic.
- Camera motion: Use MotionLoRA to guide zooms, pans, tilts and rotation.
- Image and sketch guidance: SparseCtrl accepts RGB images or sketches to control animation content alongside text prompts.
The Python implementation runs locally and includes a Gradio browser interface. Hugging Face Diffusers also supports AnimateDiff. The repository uses Apache 2.0, and the project states that it's released for academic use.
Output can show flicker, and general text-to-video quality is limited because the approach prioritizes compatibility with community image models. For image animation and interpolation, the project recommends input images generated by the same community model to keep the style consistent.
Logotyper
