PeterL1n/BackgroundMattingV2
A real-time deep learning model for high-resolution background matting of images and videos.
Not currently ranked — collecting fresh signals.
star history
This repository provides the official PyTorch implementation of the BackgroundMattingV2 paper, achieving state-of-the-art matting results at 4K 30fps and HD 60fps on consumer GPUs. The model requires a reference background image and uses a neural architecture for high-resolution refinement. It includes training code, inference scripts, and pre-trained model weights for researchers and developers to apply the matting technique.
Frequently asked
- What is PeterL1n/BackgroundMattingV2?
- A real-time deep learning model for high-resolution background matting of images and videos.
- Is BackgroundMattingV2 open source?
- Yes — PeterL1n/BackgroundMattingV2 is open source, released under the MIT license.
- What language is BackgroundMattingV2 written in?
- PeterL1n/BackgroundMattingV2 is primarily written in Python.
- How popular is BackgroundMattingV2?
- PeterL1n/BackgroundMattingV2 has 7.2k stars on GitHub.
- Where can I find BackgroundMattingV2?
- PeterL1n/BackgroundMattingV2 is on GitHub at https://github.com/PeterL1n/BackgroundMattingV2.