Human and Virtual Agent Expressive Gesture Quality Analysis and Synthesis - Coverbal Synchrony in Human-Machine Interaction - page 270

2010). Some authors also suggest that gesture's expressivity could be
important in the communication of social states such as empathy (Varni
et al., 2009) or even sexual interest (Grammer et al., 2000).
In this chapter, the expression “ expressive gesture quality” refers
to those features of nonverbal behaviors that describe how a specific
gesture is performed, for example its temporal dynamics, fluidity or
energy. In the domain of Human-Computer Interaction (HCI) , expressive
gesture quality is important at least for two reasons. First of all,
researchers try to detect the expressive qualities in human nonverbal
behavior and to infer their communicative meaning. In this case, the
final goal is the recognition of, for example, the user's emotional
state or mood. Second, expressive gesture synthesis is studied in the
design and implementation of virtual agents, i.e., anthropomorphic
autonomous characters displayed, for instance, on the computer screen
that use various verbal and nonverbal forms of communication (Cassell
et al., 2000). Virtual agents may modulate their expressive gesture
quality to better transmit their communicative intentions, e.g., their
emotional state or mood, to the user.
This chapter presents an overview of studies that enhance the
communicative capabilities of human-computer interfaces by taking
into account the expressive qualities of nonverbal behavior. It also
presents a detailed description of two systems for expressive gesture
quality analysis and synthesis in HCI. In more details, the next section
is split into three parts: in the first and second parts, we review some
of the methods described in the literature for expressive gesture quality
analysis and synthesis by illustrating various algorithms; in the third
part, we present systems in which a continuous expressive gesture
quality analysis and synthesis loop is performed. In Section 3, we
present a case study—the analysis and synthesis of some expressive
gesture features in the EyesWeb XMI platform for the creation of
multimodal applications (Camurri et al., 2007) and the virtual agent
called Greta (Niewiadomski et al., 2011). Finally, in Section 4, we
provide a conclusive overview by comparing existing algorithms for
gesture quality analysis and synthesis in HCI.
2. State of the Art
2.1 Expressive gesture quality analysis
In HCI, a central role is played by automated gesture analysis
techniques, aiming to extract and describe physical features of human
behavior and use them to infer information related to, for example,