Attach actions to an object: an object says where a person meets it
through its affordances (a seat, a handle, a surface), and this gives the
actions what they read of it, placed in the world.
The object is any model that declares affordances in its own space, such
as an archetype; nothing here depends on how it was made. Its place is
where the scene puts it. A sitting action then sits on the seat the chair
declares, at the height and depth the chair has, and a grasp, a touch or
a look goes to the contact the object names instead of to coordinates
written by hand.
Attach actions to an object: an object says where a person meets it through its affordances (a seat, a handle, a surface), and this gives the actions what they read of it, placed in the world.
The object is any model that declares affordances in its own space, such as an archetype; nothing here depends on how it was made. Its place is where the scene puts it. A sitting action then sits on the seat the chair declares, at the height and depth the chair has, and a grasp, a touch or a look goes to the contact the object names instead of to coordinates written by hand.