interactee

To that end, we explore ways to learn the generic, action-independent connections between (a) representations of a person's pose, gaze, and scene cues and (b) the interactee object's position and scale.