Detecting emergent leader in a meeting environment using nonverbal visual features only

Abstract

In this paper, we propose an effective method for emergent leader detection in meeting environments which is based on nonverbal visual features. Identifying emergent leader is an important issue for organizations. It is also a wellinvestigated topic in social psychology while a relatively new problem in social signal processing (SSP). The effectiveness of nonverbal features have been shown by many previous SSP studies. In general, the nonverbal video-based features were not more effective compared to audio-based features although, their fusion generally improved the overall performance. However, in absence of audio sensors, the accurate detection of social interactions is still crucial. Motivating from that, we propose novel, automatically extracted, nonverbal features to identify the emergent leadership. The extracted nonverbal features were based on automatically estimated visual focus of attention which is based on head pose. The evaluation of the proposed method and the defined features were realized using a new dataset which is firstly introduced in this paper including its design, collection and annotation. The effectiveness of the features and the method were also compared with many state of the art features and methods.

How attention gates social interactions

“…For example, it has been theorized that the dynamics of looking behaviors during real‐life interactions represent a tacit marker of group social processes, as both nonhuman primates and human observers viewing video recordings of group interactions look more at high‐status individuals (i.e., leaders) relative to low‐status individuals (i.e., nonleaders). Capozzi and colleagues tested this intuition during real‐life group interactions and found that group leaders could be reliably identified based on the dynamics of group‐looking behaviors, as they were looked at more and engaged in more mutual gaze episodes relative to other group members . Extending this work, Capozzi and Ristic recently found that such implicit social dynamics reliably influenced subsequent social behaviors, with the amount of time one was being looked at during a group interaction predicting the magnitude of gaze following elicited by their gaze cues in subsequent encounters.…”

“…In contrast, small groups, typically defined as those composed of 3–5 individuals engaged in a common activity, elicit and afford social interactions that require efficient handling of individual members’ social cues . Here, the three routes of processing appear to allow for the dynamic allocation of social attention to each group member depending on their identity, status, and/or situational context . In this paper, we identified the likely mechanisms that may facilitate the behavioral and processing flexibilities that are required by multiagent contexts, with further research needed to build a more consolidated understanding of how attentional operations facilitate social interactions in large and small groups.…”

Emergent Leadership Detection Across Datasets

“…Even without formal authority, emergent leaders are important for group performance [11,17], and as a result automatic identification of emergent leaders in group interactions is potentially beneficial in organisational research, for hiring decisions in the context of assessment centres [15], or for robots and intelligent agents that are supposed to interact with a group naturally. Consequently, the detection of emergent leaders is a growing topic in social signal processing [6,12,25]. These studies used nonverbal behaviour to detect emergent leaders in group interactions, which is supported by a large body of work on the connection between emergent leadership and nonverbal behaviour [1,14,16].…”

“…The PAVIS dataset [6] consists of meetings of four people each either performing a "winter survival task" or a "desert survival task". Research on the dataset focussed on detecting emergent leaders from nonverbal features only [6], using multiple kernel learning [4], or using body pose based features [7].…”

“…The PAVIS dataset [6] consists of meetings of four people each either performing a "winter survival task" or a "desert survival task". Research on the dataset focussed on detecting emergent leaders from nonverbal features only [6], using multiple kernel learning [4], or using body pose based features [7]. Further studies improved emergent leadership detection on the PAVIS dataset by using deep visual activity features [9], or by employing sequential analysis [8].…”