BiRefNet_HR-GGUF

GGUF models for BiRefNet

This BiRefNet was trained with images in 2048x2048 for higher resolution inference.

BiRefNet is a model for dichotomous image segmentation (background removal). The weights in this repository are converted for lightweight inference on consumer hardware with vision.cpp.

• Original repository: ZhengPeng7/BiRefNet (Github)
• Original weights: ZhengPeng7/BiRefNet_HR (HuggingFace)

Run

Example inference with vision.cpp:

CLI

vision-cli birefnet -m BiRefNet_HR-F16.gguf -i input.png -o mask.png --composite comp.png

C++

image_data image      = image_load("input.png");
backend_device device = backend_init();
birefnet_model model  = birefnet_load_model("BiRefNet_HR-F16.gguf", device);
image_data mask       = birefnet_compute(model, image);
image_save(mask, "mask.png");