veeliks/invoke_birefnet

Custom node for InvokeAI that provides background removal using BiRefNet

4

stars

3

commits

Python

primary language

Oct 5, 2025

updated

README

Invoke BiRefNet

A custom node for InvokeAI that provides background removal using BiRefNet (Bilateral Reference Network), a high-quality segmentation model that separates foreground objects from their backgrounds.


Portrait of a woman Portrait of a woman after background removal

Example of the node used in a workflow

Installation

  1. Copy or clone this repository into the nodes folder found in your Invoke install location
  2. Install the following requirements through the launcher console:
uv pip install timm kornia
  1. Restart Invoke and the node will appear in the image category under the name Remove Background (BiRefNet).

Notes

Models are downloaded from Hugging Face on first use and cached locally. The initial run with a new model variant will take slightly longer while it downloads.

Contributors

veeliks

3 commits

veeliks/invoke_birefnet

Custom node for InvokeAI that provides background removal using BiRefNet

4

stars

3

commits

Python

primary language

Oct 5, 2025

updated

README

Invoke BiRefNet

A custom node for InvokeAI that provides background removal using BiRefNet (Bilateral Reference Network), a high-quality segmentation model that separates foreground objects from their backgrounds.


Portrait of a woman Portrait of a woman after background removal

Example of the node used in a workflow

Installation

  1. Copy or clone this repository into the nodes folder found in your Invoke install location
  2. Install the following requirements through the launcher console:
uv pip install timm kornia
  1. Restart Invoke and the node will appear in the image category under the name Remove Background (BiRefNet).

Notes

Models are downloaded from Hugging Face on first use and cached locally. The initial run with a new model variant will take slightly longer while it downloads.

Contributors

veeliks

3 commits

Languages

Python

100.0%