Fine-grained Affordance Annotation for Egocentric Hand-Object Interaction Videos
This work addresses the need for clearer affordance definitions in hand-object interaction datasets, benefiting tasks like action anticipation and robot imitation learning, but it is incremental as it refines existing annotation approaches.
The paper tackled the problem of ambiguous affordance definitions in existing datasets by proposing an efficient annotation scheme that combines goal-irrelevant motor actions and grasp types, applied to the EPIC-KITCHENS dataset. The results showed that models trained with this annotation could distinguish affordance from other concepts, predict fine-grained interaction possibilities, and generalize across domains.
Object affordance is an important concept in hand-object interaction, providing information on action possibilities based on human motor capacity and objects' physical property thus benefiting tasks such as action anticipation and robot imitation learning. However, the definition of affordance in existing datasets often: 1) mix up affordance with object functionality; 2) confuse affordance with goal-related action; and 3) ignore human motor capacity. This paper proposes an efficient annotation scheme to address these issues by combining goal-irrelevant motor actions and grasp types as affordance labels and introducing the concept of mechanical action to represent the action possibilities between two objects. We provide new annotations by applying this scheme to the EPIC-KITCHENS dataset and test our annotation with tasks such as affordance recognition, hand-object interaction hotspots prediction, and cross-domain evaluation of affordance. The results show that models trained with our annotation can distinguish affordance from other concepts, predict fine-grained interaction possibilities on objects, and generalize through different domains.