Author:

Chen, Vanessa
Supervisor:Prof. Gudrun Klinker
Advisor:Eghtebas, Nassim (@ga53xoy)
Submission Date:[created]

Abstract

Although synchrony in physical conversations through facial expressions and body movements is thoroughly researched, little is known about whether synchrony can be detected or even measured in conversations held virtually. This research aims to analyze visually captured facial cues in virtual conversations and to investigate if synchrony can be observed. Therefore, we used a dataset of 10 virtual conversations where strangers talked in 5 to 10 minutes intervals on 3 different topics. These topics were an ice-breaker and two given corporate tasks, which the dyads had to complete. Several methods were explored to visually detect heart rates, smiles and eye blinks synchrony, and a sentiment analysis was conducted to examine contextual synchrony. Not enough reliable heart rate values were received to identify synchrony between the conversation partners. However, a wavelet coherence analysis on the eye blinks and smile detection indicates some synchrony. This exploration of facial cues in virtual conferences extends our knowledge of conversational synchrony.

Results/Implementation/Project Description

Conclusion

[ PDF (optional) ] 

[ Slides Kickoff/Final (optional)]