ADetailer is an extension for the Stable Diffusion WebUI that performs automatic detection, masking, and inpainting.
The tool addresses the need to selectively refine specific regions of generated images without manual mask creation. It works by using detection models to automatically identify objects of interest, generate masks around them, and then apply inpainting to those masked areas. Users can configure which object classes to detect, set confidence thresholds to filter detections, and control mask properties through erosion, dilation, and offset parameters before inpainting is applied.
ADetailer suits workflows where you want to enhance particular elements in an image—such as faces, hands, or other specific objects—without affecting the rest of the composition. It integrates directly into the WebUI as an extension and can be installed from the Extensions tab or via git URL. The tool supports multiple detection models, including YOLO World models that allow custom class specification, and provides fine-grained control over mask preprocessing through offset, erosion/dilation, and merge modes. It is positioned as similar to Detection Detailer but offers its own implementation within the WebUI ecosystem.
Development activity shows consistent engagement with the codebase through regular updates and refinements to detection and masking logic. The project maintains active responsiveness to issues and pull requests, indicating ongoing maintenance and community interaction. Documentation is thorough, with detailed option tables and configuration guidance provided in the README to support users in tuning detection and inpainting parameters.
