Seedream V5.0 Flash Layer Decomposition
Seedream V5.0 Flash Layer Decomposition turns flattened artwork into separately editable components. It takes a source image and produces a base plus transparent layers that editors can use in a compositor or design application.
Optional text guidance directs which objects to isolate and how elements should be grouped. With image-only input, the model identifies important visual elements automatically. This makes it useful both for targeted extraction and for breaking down an image into components for further editing.
What it's good for
Designers can prepare campaign artwork for layout revisions by separating products or foreground subjects from surrounding imagery. The extracted elements become reusable material for other compositions, and individual layers can receive retouching without requiring the entire image to be handled as one piece.
Compositors can use the outputs to rearrange scene elements or replace isolated content during subsequent editing. This is particularly useful when the available source is a flattened image.
