In this paper, we have investigated the comparative usability among three different viewing configurations of augmented reality (AR) system that uses a desktop monitor instead of a head mounted display. In many cases, due to operational or cost reasons, the use of head mounted displays may not be viable. Such a configuration is bound to cause usability problems because of the mismatch in the user's proprioception, scale, hand eye coordination, and the reduced 3D depth perception. We asked a pool of subjects to carry out an object manipulation task in three different desktop AR set ups. We measured the subject's task performance and surveyed for the perceived usability and preference. Our results indicated that placing a fixed camera in the back of the user was the best option for convenience and attaching a camera on the user�s head for task performance. The results should provide a valuable guide for designing desktop augmented reality systems without head mounted displays
Virtual exertions: A user interface combining visual information, kinesthetics and biofeedback for virtual object manipulation
