A User Perception--Based Approach to Create Smiling Embodied Conversational Agents

Abstract

In order to improve the social capabilities of embodied conversational agents, we propose a computational model to enable agents to automatically select and display appropriate smiling behavior during humanmachine interaction. A smile may convey different communicative intentions depending on subtle characteristics of the facial expression and contextual cues. So, to construct such a model, as a first step, we explore the morphological and dynamic characteristics of different types of smile (polite, amused and embarrassed smiles) that an embodied conversational agent may display. The resulting lexicon of smiles is based on a corpus of virtual agent's smiles directly created by users and analyzed through a machine learning technique. Moreover, during an interaction, the expression of smile impacts on the observer's perception of the interpersonal stance of the speaker. As a second step, we propose a probabilistic model to automatically compute the user's potential perception of the embodied conversational agent's social stance depending on its smiling behavior and on its physical appearance. This model, based on a corpus of users' perception of smiling and non-smiling virtual agents, enables a virtual agent to determine the appropriate smiling behavior to adopt given the interpersonal stance it wants to express. An experiment using real human-virtual agent interaction provided some validation of the proposed model.

Anthropomorphic Information Systems

“…One of the primary goals of anthropomorphic design is to positively influence humans' affect, which is observed to be an important factor in human-computer interaction (HCI) (Hudlicka 2003;Ochs et al 2017). Imbuing systems or robots with anthropomorphic cues may have many positive implications, such as higher likeability due to social or emotional connection (Eyssel et al 2010), an increase of decision quality in purchase processes because of a more natural interaction (Qiu and Benbasat 2009) or even improvements in sociability of children diagnosed with autism spectrum conditions (Bernardini et al 2014).…”

Classifying Smart Personal Assistants: An Empirical Cluster Analysis

“…Embodied or anthropomorphic design is usually applied to provide a shared common ground, represent an authentic entity, combine verbal and non-verbal communication and align minds by being interesting, creative and humorous [22]. In practice, embodiment is accomplished by virtual characters, i.e., avatars [10,24], a (often human-like) computer voice [36] or a combination of both [44]. However, some SPAs do not use embodiment at all [38].…”

Time to Go ONLINE! A Modular Framework for Building Internet-based Socially Interactive Agents

“…Indeed, social appropriateness during dialogues, requires (apart from choosing an appropriate topic) knowing how to use different channels of communication to establish and maintain rapport via verbal and non-verbal cues [9,22], such as respectful eye contact [10], motor mimicry and synchoronous postures [3], expression of facial and other nonverbal social cues that are congruent with verbal utterances and emotional states, among others. In spite of such complexity, IVA researchers have leveraged latest progress in affective computing [2,5,6,8] to build agents with social cues and responses [1,10,13,16,18,18,24] towards becoming able to establish some rapport [11] and express empathy [12,14,17].…”