Abstract: Scene Graph Generation (SGG) aims to parse images into structured graph representations composed of objects and their relationships. Despite recent advances, SGG remains faced with two ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results