bytedance/X-Dyna
[CVPR 2025 Highlight] X-Dyna: Expressive Dynamic Human Image Animation
This project helps create realistic, expressive video animations of people from just a single image. You provide an image of a person and a separate video demonstrating the desired body movements and facial expressions. The output is a new video of the person from your image performing those actions with dynamic, lifelike details, ideal for content creators, animators, or marketing professionals.
264 stars. No commits in the last 6 months.
Use this if you need to animate a static human image with complex, natural body movements and nuanced facial expressions from a driving video, ensuring dynamic and context-aware scene details.
Not ideal if you lack access to a powerful GPU (minimum 20GB memory) or only need simple, non-expressive pose transfers without dynamic scene elements.
Stars
264
Forks
24
Language
Python
License
Apache-2.0
Category
Last pushed
Jan 30, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/bytedance/X-Dyna"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
hao-ai-lab/FastVideo
A unified inference and post-training framework for accelerated video generation.
ModelTC/LightX2V
Light Image Video Generation Inference Framework
thu-ml/TurboDiffusion
TurboDiffusion: 100–200× Acceleration for Video Diffusion Models
PKU-YuanGroup/Helios
Helios: Real Real-Time Long Video Generation Model
PKU-YuanGroup/MagicTime
[TPAMI 2025🔥] MagicTime: Time-lapse Video Generation Models as Metamorphic Simulators