Towards Realism: Selection Techniques for Virtual Human Interaction

Abstract

Virtual community simulation has been engaged on extensive research studies with rapid growth of virtual worlds. Realistic interaction with virtual humans is complex open research area in virtual community simulation, where significant effort has been devoted to simulate natural phenomena. This research explores a natural user interface for interacting with virtual humans under selection term. By combining and extending known virtual object selection techniques, this research introduces two head oriented selection techniques which are called voice-ray casting metaphor and voice casting metaphor. Selection strategies of these metaphors are defined to overcome present limitations in existing techniques such as memory overhead and cognitive effort over the user. Additionally these metaphors are consisted with a behaviour model which simulates neighbourhood behaviour changes of virtual crowd in order to bring a sense of reality in the simulation. System prototype is integrated with multiple modalities such as head mounted display, joysticks and headsets which provides an immersive user experience in human computer interaction. A user study has been conducted to comparatively evaluate proposed selection techniques with ray casting metaphor which is a widely used metaphor in games and simulations. The experimentation is conducted for evaluating in both ideal and walking states of virtual humans. The metaphor evaluation is based on the defined parameters of interactive experiment such as simplicity, intuitiveness and realism. The user study results show suitability of proposed head oriented metaphors for virtual community simulations and the challenges to be improved in the domain of virtual human interaction.