TMElyralab/MusePose

Health v2 · maintenance only

Flags: no_releases no_license

Adoption not part of the score

2701 stars · 205 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

MusePose is a diffusion-based, pose-guided image-to-video generation framework for creating virtual human videos, where a character in a reference image is animated according to a given pose sequence. It is part of Tencent Music's Muse open-source series alongside MuseV and MuseTalk, and includes training code and a pose alignment algorithm.

Use cases

• generate dance videos from a reference image and pose sequence
• animate a virtual human character from a single photo
• align arbitrary dance videos to reference images for better inference
• train a pose-driven video generation model
• build end-to-end virtual human generation pipelines
• reproduce or extend AnimateAnyone research

When to choose

• you need pose-controlled human video generation from a still image
• you want an open-source implementation of AnimateAnyone with training code
• you are building virtual human or digital avatar pipelines
• you need high-quality dance video synthesis from open models

When to avoid

• you need general-purpose text-to-video without pose control
• you lack a GPU or cannot run diffusion models locally
• you need lip-sync or talking-head generation alone (use MuseTalk instead)
• you require a permissively licensed model for commercial use without checking the custom license

Facets

framework · maturity active

machine-learning deep-learning video-processing image-processing llm-training deep-learning computer-vision artificial-intelligence image-processing python image-to-video pose-guided-generation diffusion-models virtual-human dance-generation animateanyone video linux gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="TMElyralab/MusePose")