This repository provides a comprehensive suite of ComfyUI nodes designed to replicate Photoshop-like layer and mask compositing functionalities. It aims to streamline creative workflows by centralizing image manipulation tasks within ComfyUI, reducing the need to switch between different software applications. The target audience includes digital artists, graphic designers, and AI art enthusiasts who utilize ComfyUI for image generation and editing.

How It Works

The project offers a wide array of nodes categorized into LayerStyle, LayerColor, LayerMask, LayerUtility, and LayerFilter. These nodes implement various image manipulation techniques, including layer blending, color adjustments, mask generation, text rendering, and image effects. Many nodes leverage advanced AI models for tasks like segmentation, background removal, and image-to-prompt generation, enabling sophisticated image processing directly within the ComfyUI environment.

Quick Start & Requirements

• Installation: Recommended via ComfyUI Manager. Alternatively, clone the repository into the custom_nodes directory and run install_requirements.bat (or install_requirements_aki.bat for Aki ComfyUI).
• Dependencies: Requires Python and specific libraries installed via requirements.txt. Model files need to be downloaded separately from Hugging Face or BaiduNetdisk and placed in the ComfyUI/models folder.
• Resources: Some "Ultra" nodes and models like LlamaVision require significant VRAM (e.g., 16GB+ for PhiPrompt).

Highlighted Details

• Extensive Photoshop-like layer styles (Drop Shadow, Outer Glow, Inner Shadow, Stroke, etc.).
• Advanced AI-powered segmentation and background removal nodes (e.g., SAM2Ultra, RemBgUltra).
• Numerous utility nodes for image manipulation, text generation, and workflow debugging.
• Support for various AI models including Florence2, SAM2, YOLO, and more.
• Frequent updates and additions of new nodes and features.

Maintenance & Community

The project is actively maintained with frequent commits and pull request merges. Community support and contributions are encouraged, with specific acknowledgments for code contributions. Contact information for business cooperation is provided.

Licensing & Compatibility

The LayerStyle nodes are MIT licensed. However, the README notes that some functional code originates from other open-source projects, and commercial use may require authorization agreements from those original projects.

Limitations & Caveats

• A significant number of advanced nodes have been split into a separate repository (ComfyUI_LayerStyle_Advance), requiring users to install both for full functionality.
• Some nodes have specific, high VRAM requirements.
• Dependency management can be complex, with specific instructions for resolving common errors related to package versions (e.g., opencv-contrib-python, transformers).
• Users may need to manually download model files for many nodes.