Bilkent University
Department of Computer Engineering


Spatiotemporal and Multimodal Analysis of Personality Traits


Dersu Giritlioğlu
MS Student
Computer Engineering Department
Bilkent University

Personality traits are being studied by a wide range of scientific fields like psychology, psychiatry and neuroscience. There is an increasing interest in computer science area towards this topic with the help of enhanced computation power. We are currently able to detect visual cues such as, expressions of the human face, gaze, posture and gestures to analyze visible personality traits. But these cues are not ready to be used as reliable, fast and accurate assessment tools and we cannot automatize personality trait evaluation yet. In this study, we present a spatio-temporal and multimodal approach using audio-visual and transcribed information to analyze Big Five personality traits (OCEAN). Along with the present datasets in use, we created out own database, which is collected via our own robust desktop application that automatizes the data collection process.


DATE: 09 December 2019, Monday @ 16:40