ChainHOI: Joint-based Kinematic Chain Modeling for Human-Object Interaction Generation

Someone is shoving the suitcase with a kick. A person is grasping a trashbin. Someone is sitting in a chairblack. A person is toting a backpack. A person is grasping the boxmedium from the front. Figure 1 . Given a text description and target object geometry, our ChainHOI effectively generates high-quality human-object interaction sequences that are both logical and realistic.