This extension for Stable Diffusion WebUI, BMAB, automates post-processing of generated images. It targets users looking to refine AI-generated visuals, particularly those featuring people, by enabling automatic masking and inpainting of faces and hands, alongside image resizing and compositing.

How It Works

BMAB leverages detection models to identify and process specific image regions, primarily focusing on human anatomy like faces and hands. It supports a pipeline of operations including inpainting, resizing, resampling, noise addition, and image compositing. This modular approach allows for complex post-processing workflows to be applied automatically, enhancing the quality and consistency of generated images.

Quick Start & Requirements

• Install as a Stable Diffusion WebUI extension.
• Requires Stable Diffusion WebUI.
• Manual (KR): https://www.facebook.com/profile.php?id=61553793992101

Highlighted Details

• Automatic masking and inpainting for person, face, and hand.
• Image resizing using detection models.
• Supports compositing two images and upscaling.
• Integrates ControlNet features like Noise and Pose.

Maintenance & Community

• Primarily developed by portu-sim.
• Community engagement appears limited, with a Facebook link provided for examples.

Licensing & Compatibility

• License not specified in the README.
• Compatibility is tied to the Stable Diffusion WebUI ecosystem.

Limitations & Caveats

The README provides extensive examples of parameter combinations but lacks explicit installation instructions or details on underlying model dependencies. The primary community link is a Facebook profile, suggesting a potentially small or informal development community.