About this workflow
A compact MOSS Image workflow demonstrating localized image generation and editing using BBoxes (Bounding Boxes).
Instead of describing the entire image in a single prompt, you can divide the composition into separate regions and assign each one its own instructions. This allows you to control individual elements independently, such as characters, clothing, logos, headlines, typography, graphic elements, and other specific details.
It's particularly useful for creating magazine covers, movie posters, promotional artwork, character designs, and other compositions that require precise control over multiple visual elements.
Included: Civitai Crybaby
This workflow also includes my Civitai Crybaby custom node, which provides prompt scanning and notifications to help identify potentially problematic prompt terms before generation.
Installation: https://civarchive.com/models/3025394/civitai-crybaby
Essential : Enable Nodes 2.0
You must enable Nodes 2.0 in ComfyUI to use the BBox functionality.
To enable it:
Click the ComfyUI logo in the top-left corner.
Find and enable the Nodes 2.0 option in the menu.
Without this feature enabled, you won't be able to interact with the BBox workflow.
Have fun!