AffectiveSpotlight: Facilitating the Communication of Affective Responses from Audience Members during Online Presentations

Abstract

Figure 1: Existing videoconferencing platforms such as Microsoft Teams provide limited audience feedback when giving presentations (left). We develop AffectiveSpotlight which analyzes and spotlights audience members in real-time to support presenters (right). Text boxes indicate the main components of the platforms.

I Cannot See Students Focusing on My Presentation; Are They Following Me? Continuous Monitoring of Student Engagement through "Stungage"

“…Besides the metric-based analysis, these experiments were evaluated using a set of surveys consisting of system evaluation and student understandability. Once the class is over, three different layouts are shared with the instructor, and for each layout, they were asked to fill up (1) system evaluation survey [36,37] that captures the instructor's assessment on the system, and (2) student understandability survey [36,38] that captures instructor's experience with the view. We perform Paired Wilcoxon signed-rank tests with correction on all the survey questions to understand the differences in the survey responses across the different layouts.…”

CalmResponses: Displaying Collective Audience Reactions in Remote Communication

“…To obtain real-time reactions of multiple audiences in online situations unobtrusively, we require a scalable system that only uses commonly available devices, such as laptops. In this sense, AffectiveSpotlight [50] is a suitable system to share audience reactions with speakers because it estimates facial responses and head gestures of audiences with a webcam. However, it presents only one audience with the highest engagement simultaneously and it is impossible for speakers to grasp how the others feel because it is hard to deny that while one audience is engaging, the others are not during presentations.…”

“…In each experiment, we performed three evaluations of the system: objective and subjective evaluation of the presentations, number of fillers, and qualitative comments. These are widely used measures in existing works to evaluate speech quality and indirectly estimate speakers' states such as anxiety [50,66].…”

“…This method can obtain audience reactions without burden. Existing research has exploited these signals as audience reactions [20,40,50,62].…”

Bridging Social Distance During Social Distancing: Exploring Social Talk and Remote Collegiality in Video Conferencing

“…Indeed, specific functionality could also be added to VC platforms to encourage and equalize participation and engagement in social talk across teams. Similar to recent work on assessing meeting dynamics (Samrose et al, 2021) and spotlighting audience reactions in VC meetings (Murali et al, 2021), a bot could be embedded in social talk VC sessions to read conversational dynamics or highlight specific reactions so as to encourage more social talk and support those that have not contributed within a session. This type of functionality could be useful to highlight opportunities for engagement in group conversations and encourage a wider range of contributions from group members.…”