AliaksandrSiarohin/first-order-model

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

15015 stars · 3256 forks observed · 2026-08-28

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

Official PyTorch/Jupyter implementation of the First Order Motion Model for image animation (NeurIPS 2019). It animates a static source image (e.g., a face or body) using motion transferred from a driving video, without object-specific annotations.

Use cases

• animate a photo with motion from a driving video
• make a face talk using a video of someone else
• transfer facial expressions from video to a still image
• create deepfake-style talking head videos
• retarget motion from one person to another
• reproduce the First Order Motion Model paper results

When to choose

• you want to animate a still image from a driving video using a proven research model
• you need pre-trained checkpoints for faces (VoxCeleb), fashion, or animated graphics
• you want a reproducible implementation of the NeurIPS 2019 paper

When to avoid

• you need a production-ready or maintained product rather than research code
• you need articulated-object animation - the authors recommend their newer articulated-animation framework
• you need real-time animation or a polished user interface

Facets

library · maturity maintenance

deep-learning image-processing video-processing machine-learning deep-learning computer-vision image-processing python image-animation motion-transfer deepfake generative-model face-animation neurips-2019 research-code video linux gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="AliaksandrSiarohin/first-order-model")