Empirical evaluation and pathway modeling of visual attention to virtual humans in an appearance fidelity continuum

• Published in: Journal on multimodal user interfaces Vol. 15; no. 2; pp. 109 - 119
• Main Authors: Volonte, Matias, Anaraky, Reza Ghaiumy, Venkatakrishnan, Rohith, Venkatakrishnan, Roshan, Knijnenburg, Bart P., Duchowski, Andrew T., Babu, Sabarish V.
• Format: Journal Article
• Language: English
• Published: Cham Springer International Publishing 01.06.2021; Springer Nature B.V
• Subjects: Computer Science; Empirical analysis; Eye movements; Image Processing and Computer Vision; Original Paper; Rendering; Signal,Image and Speech Processing; Tracking systems; Training simulators; User Interfaces and Human Computer Interaction; Virtual humans; Human computer interaction; Virtual Agents; Eye tracking; Virtual training simulator; Rendering style; Virtual humans
• ISSN: 1783-7677; 1783-8738
• DOI: 10.1007/s12193-020-00341-z

In this contribution we studied how different rendering styles of a virtual human impacted users’ visual attention in an interactive medical training simulator. In a mixed design experiment, 78 participants interacted with a virtual human representing a sample from the non-photorealistic (NPR) to the photorealistic (PR) rendering continuity. We presented five rendering style samples scenarios, namely All Pencil Shaded (APS), Pencil Shaded (PS), All Cartoon Shaded (ACT), Cartoon Shaded (CT), and Human-Like (HL), and compared how visual attention differed between groups of users. For this study, we employed an eye tracking system for collecting and analyzing users’ gaze during interaction with the virtual human in a failure to rescue medical training simulation. Results shows that users spent more total time in the APS and ACT conditions but users visually attended more to virtual humans in the PS, CT and HL appearance conditions.