This project implements Robust Video Matting (RVM) inside The Foundry's Nuke, enabling high-quality real-time background matting using deep learning.
This adaptation allows RVM to run natively within Nuke as a Cattery model, making it easy to integrate into existing compositing workflows without requiring additional dependencies.
- High-quality real-time video matting with detail preservation
- Optimized deep learning model using
rvm_mobilenetv3.pth - Support for RGB and Alpha channels, enabling direct compositing
- Works on a wide range of backgrounds without requiring trimaps
- Lightweight and efficient, allowing fast inference directly in Nuke
- Tested on Nuke 14.1v5 and Linux
- Download and extract the latest release.
- Copy the extracted
Catteryfolder to your.nukedirectory or Nuke's plugin path. - In Nuke, go to Cattery > Update, or restart Nuke.
- The RVM node will be accessible under Cattery > Matting > RVM.
This implementation uses the MobileNetV3-based RVM model (rvm_mobilenetv3.pth), which is optimized for real-time performance while maintaining matting results.
For more details on the original model, visit: Robust Video Matting GitHub
- Version: 1.0.1
- Release Date: Feb.07.2025
- Developer: Dean Kwon (deankwon724@gmail.com)
This project is adapted from PeterL1n/RobustVideoMatting and follows the original licensing terms.
Users are responsible for ensuring compliance with licensing requirements when using the model and its dependencies.