magic-research/magic-animate

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

10897 stars · 1081 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

MagicAnimate is the official implementation of a CVPR 2024 diffusion-based human image animation framework that animates a reference image following motion sequences from a driving video with temporal consistency. It builds on Stable Diffusion 1.5 with a DensePose-based ControlNet, appearance encoder, and temporal attention modules, and ships with inference code and a Gradio demo.

Use cases

• animate a person photo using motion from a video
• transfer dance moves from a video onto a reference image
• animate oil paintings or movie characters performing actions
• animate multiple people following a given motion sequence
• combine text-to-image generation with human animation
• cross-identity animation using motion sequences from different people

When to choose

• you need research-grade human image animation with temporal consistency
• you want to animate stylized or out-of-domain images like paintings
• you want a reproducible CVPR 2024 baseline for pose-driven video generation
• you have GPU resources and want local inference or a Gradio demo

When to avoid

• you need real-time animation on consumer hardware
• you want a polished end-user product rather than research code
• you cannot download multi-gigabyte Stable Diffusion checkpoints
• you need active feature development or commercial support

Facets

library · maturity maintenance

video-processing image-processing machine-learning deep-learning computer-vision image-processing artificial-intelligence deep-learning python diffusion-model human-animation pose-transfer stable-diffusion temporal-consistency cvpr-2024 gradio-demo video gpu linux

Member repositories

For agents

markdown · JSON · MCP: product_card(name="magic-research/magic-animate")