Analyzing Group Interactions in Conversations: a Review

Abstract

Abstract-Multiparty face-to-face conversations in professional and social settings represent an emerging research domain for which automatic activity-based analysis is relevant for scientific and practical reasons. The activity patterns emerging from groups engaged in conversations are intrinsically multimodal and thus constitute interesting target problems for multistream and multisensor fusion techniques. In this paper, a summarized review of the literature on automatic analysis of group activities in face-to-face conversational settings is presented. A basic categorization of group activities is proposed based on their typical temporal scale, and existing works are then discussed for various types of activities and trends including addressing, turn taking, interest, and dominance.

Social Behavior Modeling Based on Incremental Discrete Hidden Markov Models

“…First, speaker diarization is applied to list turns; second, acoustic features are extracted from turns and finally, features vectors are mapped into a sequence of roles. More complete reviews on issues and models related to nonverbal analysis of social interaction can be found in [11] [7]. In the context of multimodal behavior generation, several platforms have been proposed for humanoid robots and virtual agents.…”

“…This model should thus be able to perceive and understand the partner's actions on their joint environment and generate adequate actions that should reflect its current understanding of the evolution of the joint plan. A complete interaction can be seen as a sequence of discrete tasks, sub-tasks or activities [11]. In the following, we will consider a situated conversation as a sequence of cognitive states that structure the joint behaviors of the conversation partners.…”

Analyzing Perceived Empathy Based on Reaction Time in Behavioral Mimicry

“…To realize such applications, it's required to automatically understand not only human behavior but also the participants' emotions which temporally evolve in the course of the interaction and impacts the conversation. Accordingly, the main target of automatic meeting analysis is now shifting from behavior to emotion [1]- [5].…”

Emergent leaders through looking and speaking: from audio-visual data to multimodal recognition

“…Psychologists have long studied these interactions of varying scale, to understand behavior, motivation, and emergence of interaction patterns (Poole et al, 2004;Salas et al, 2005). One specific setting of importance is the study of small groups (Bales and Strodtbeck, 1951;Gatica-Perez, 2006).…”