Ideal inputs
Upload one scene image where the unwanted object is visible and describe its location plus the lines, surface, or background that should continue behind it.
Remove a described object from a photo and rebuild the area with surrounding texture, perspective, and light.
Upload one scene image where the unwanted object is visible and describe its location plus the lines, surface, or background that should continue behind it.
Choose the clearest source and keep the target object fully visible in the frame.
Name the object, position, approximate size, and nearby surfaces that should be reconstructed.
Check shadows, repeated patterns, edges, reflections, and any subject parts close to the removed object.
The edit workflow creates a cleaned scene concept with the described object removed where the surrounding context supports reconstruction.
There is no dedicated mask editor, and large or structurally important objects can leave visible artifacts. Verify the filled area before publishing.
Use a plain description and include its position, size, and nearby texture so the model can distinguish it from objects that should stay.
It can attempt the edit, but content hidden behind an object is not known. Review faces and important details carefully.