Skip to content

DisPose

Animate human images with motion field guidance

Visit official site

What dispose is

The method is described as a way to control human image animation, using motion field guidance and keypoint correspondence to keep motion and identity aligned.

What you can do with it

  • Download the code and read the setup docs
  • Prepare a reference image and a driving video
  • Run the code to generate a new video

What to watch out for

  • Dense guidance works at region level
  • Existing model parameters are not frozen
  • Depth maps are not listed as required input

Pros & cons

✓ What we like

  • No extra dense inputs such as depth maps are needed
  • Region-level dense guidance is provided

! What to watch out for

  • A reference image and a driving video are prepared by the user
  • Environment setup and dependencies come before running

FAQ

What inputs does it need?

The material names a reference image and a driving video.

Does it change model weights?

The material says the plug-in module improves quality without freezing existing model parameters.

Last reviewed: 2026-11-15

More AI video generation tools

View all →

How we review