What aspects of realism and faithfulness are relevant to supporting non-verbal communication through 3D mediums

Campion, SP

Abstract

This thesis investigates what aspects of realism and faithfulness are relevant to supporting non-verbal communication through visual mediums. The mediums examined are 2D video, 3D computer graphics and video based 3D reconstruction. The latter is 3D CGI derived from multiple streams of 2D video. People’s ability to identify behaviour of primates through gross non-verbal communication is compared across 2D video and 3D CGI. Findings suggest 3D CGI performs equally well to 2D video for the identification of gross non-verbal behaviour, however user feedback points to a lack of understanding of intent. Secondly, ability to detect truthfulness in humans across 2D video and video based 3D reconstruction mediums is examined. Effort of doing this is measured by studying changes in level of oxygenation to the prefrontal cortex. Discussion links to the literature to propose that tendency to over trust is inversely proportional to the range of non-verbal resources communicated through a medium. It is suggested that perhaps this is because “tells” are hidden. The third study identifies that video based 3D reconstruction can successfully illustrate subtle facial muscle movements on a par with 2D video, but does identify issues with the display of lower facial detail, due to a reconstruction error called droop. It is hoped that the combination of these strands of research will help users, and application developers, make more informed decisions when selecting which type of virtual character to implement for a particular application therefore contributing to the fields of virtual characters and virtual environments/serious gaming, by giving readers a greater understanding of virtual characters ability to convey non-verbal behaviour.

Citation

Campion, S. What aspects of realism and faithfulness are relevant to supporting non-verbal communication through 3D mediums. (Thesis). The University of Salford

• Thesis Type: Thesis
• Deposit Date: Sep 2, 2016
• Publicly Available Date: Sep 2, 2016
• Award Date: Jun 9, 2016