YBYBZhang/ControlVideo
A training-free controllable text-to-video generation system using diffusion models and ControlNet.
Not currently ranked — collecting fresh signals.
star history
ControlVideo adapts ControlNet for video generation without any fine-tuning, enabling high-quality and temporally consistent text-to-video synthesis. The system uses Stable Diffusion v1.5 as its base model combined with ControlNet conditioning mechanisms for spatial control. It incorporates frame interpolation using RIFE and offers support for both ControlNet 1.0 and 1.1 versions.
Frequently asked
- What is YBYBZhang/ControlVideo?
- A training-free controllable text-to-video generation system using diffusion models and ControlNet.
- Is ControlVideo open source?
- Yes — YBYBZhang/ControlVideo is open source, released under the MIT license.
- What language is ControlVideo written in?
- YBYBZhang/ControlVideo is primarily written in Python.
- How popular is ControlVideo?
- YBYBZhang/ControlVideo has 863 stars on GitHub.
- Where can I find ControlVideo?
- YBYBZhang/ControlVideo is on GitHub at https://github.com/YBYBZhang/ControlVideo.