Person in Place: Generating Associative Skeleton-Guidance Maps for Human-Object Interaction Image Editing