Custom node designed for ComfyUI that allows users to incorporate lllyasviel's IC-Light V2 model into their projects via the fal.ai API. This tool facilitates advanced image manipulation, including relighting and background changes based on textual prompts.
- Integrates seamlessly with ComfyUI, enhancing its capabilities with advanced lighting and background adjustments.
- Offers adjustable parameters for fine-tuning image generation, providing users with greater control over the output.
- Supports multiple image generation in a single request, streamlining workflows for users needing variations.
Context
This custom node enhances the ComfyUI framework by integrating the IC-Light V2 model from lllyasviel, which specializes in re-lighting images and altering backgrounds. Utilizing the fal.ai API, it allows users to manipulate images based on detailed descriptions, making it a powerful tool for artists and developers working with AI-generated visuals.
Key Features & Benefits
The node provides several practical features, including the ability to re-light photos and change backgrounds according to user-defined prompts. Users can also input their own ComfyUI masks, enhancing specificity in image processing. Adjustable parameters like seed, lighting direction, and denoise strength allow for precise control over the final output, ensuring that users can achieve the desired visual effects.
Advanced Functionalities
Among its advanced capabilities, the node allows for multiple image generations in a single request, which is particularly beneficial for users who require various iterations of an image. Additionally, it supports high-resolution fixes and customizable denoise strengths, enabling users to enhance image quality further. The option to use both positive and negative prompts helps guide the model in achieving more accurate results.
Practical Benefits
This tool significantly improves workflow efficiency in ComfyUI by streamlining the process of image manipulation. Users gain enhanced control over the lighting and background of their images, which can lead to higher quality outputs. By allowing for multiple variations in one request, it reduces the time spent on generating different images, making the overall process more efficient.
Credits/Acknowledgments
The development of this tool is credited to lllyasviel for the IC-Light model and comfyanonymous for the ComfyUI framework. Special thanks to behzad kazemi for the example workflow photo and fal.ai for providing the essential API endpoint. The project is licensed under the MIT License, ensuring its open-source availability for further development and use.