Analysing Listener Behaviour Through Gaze Data and User Performance during a Sound Localisation Task in a VR Environment

Virtual reality (VR) is used today in a variety of applications. Some examples include virtual meetings, simulators (driving, flight, medical interventions), digital twins and healthcare systems. In this context, audio plays a significant role in adding emotion and connecting users to what they see on the screen. For this reason, it is critical to evaluate how users perceive these stimuli in order to improve their quality of experience (QoE). The main objective of this work is to evaluate how listeners perceive spatialised audio in a VR environment. To accomplish this goal, users are required to localise multiple sound sources, while also listening to distractors, in the acoustic scene. In addition, this work investigates user interaction with the environment, by comparing two interaction methods used to localise the sound source. Findings from this experiment show that eye gaze can be associated with levels of effort and cognitive load in a sound localisation task. Furthermore, other metrics related to eye gaze such as the number of fixations and number of revisits to virtual objects were also used as descriptors of user attention in this task. The results obtained improve the understanding of a listener's behaviour in VR.