A methodological approach relating the classification of gesture to identification of human intent in the context of human-robot interaction

Abstract

In order to infer intent from gesture, a broad classification of types of gestures into five main classes is introduced. The classification is intended as a generally applicable basis for incorporating the understanding of gesture into human-robot interaction (HRI). Examples from human-robot interaction show the need to take into account not only the kinematics of gesture, but also the interactional context. Requirements for the operational classification of gesture by a robot interacting with humans are suggested and initial steps in its deployment are discussed

Cooperative gestures: Effective signaling for humanoid robots

“…From as early as 14 months of age, humans communicate cooperatively, to inform others of things and to share interest [1]. Nehaniv [2] refers to these gestures as interactional, used to "regulate interaction with a partner, i.e. used to initiate, maintain, invite, synchronize, organize or terminate a particular interactive, cooperative behavior."…”

Helping robots imitate: metrics and technological solutions inspired by human behaviour

“…Our studies also clearly suggest an interaction between the type of task and the type of gestures produced. This point stresses the importance of knowing the context in which gestures are produced and the interaction history [15]. Comparing the results from both studies, what seems more salient is the willingness of people to engage in interactions with the robot even if the trigger for such is just a 'small' acknowledgement of their actions.…”

“…The conceptual framework presented in [15] was adapted to capture requirements for contextual interpretation of body postures and human activities for purposes of HRI. It defines five functional classes of gestures:…”

“…But it is important to note that certain gestures in particular situations might be multipurpose. Nehaniv et al [15] stress the importance of knowing the context in which gestures are produced since it is crucial to disambiguate their meaning. In practice, data on the interaction history and context may help the classification process.…”

Issues in Human/Robot Task Structuring and Teaching

“…The coding scheme utilized here to classify the observed gestures is in line with the multi-functional classification system proposed by Nehaniv et al [7]. The coding scheme has been developed, validated and successfully used in other studies [see, 8,9].…”