Regional Guidance Layers

Regional guidance layers work by regenerating identified parts of an image using additional input. Additional guidance can be given in the form of positive text prompts, negative text prompts, and reference images.

Using a text-based Regional Guidance Layer

To add positive or negative text-based regional guidance, go to the Layers tab, click the ‘Add Layer’ button, and select Regional Guidance. Next, choose how you’d like to define the layer from the toolbar—using the brush tool is usually the easiest method. Select the area of your image where you want to provide guidance using the canvas drawing tools and enter the prompt guidance.

Using an image-based Regional Guidance Layer

Regional Reference Image layers use Image Prompt Adapter (IP-Adapter) models to inspire a new image with the content of an input image. You can add image-based Regional Guidance layers by going to the Layers tab, click the ‘Add Layer’ button, and selecting ‘Reference Image’ within the Regional Guidance layer menu.

You can read more about Reference Image layers here.