Our two IEEE VR 2019 manuscripts have been accepted. I will present our works in Osaka Japan in March 2019.
Our first paper is called “Do Head-Mounted Display Stereo Deficiencies Affect 3D Pointing Tasks in AR and VR?”. You can find the abstract of this paper below:
Most AR and VR headsets use stereoscopic displays to show virtual objects in 3D. However, the limitations of current stereo display systems affect depth perception through conflicting depth cues, which then also affect virtual hand interaction in peri-personal space, i.e., within arm’s reach. We performed a Fitts’ law experiment to understand better the impact of stereo display deficiencies of AR and VR headsets on pointing at close-by targets arranged laterally or along the line of sight. According to our results, the movement direction and the corresponding change in target depth affect pointing time and throughput; subjects’ movements towards/away from their head were slower and less accurate than their lateral movements (left/right). However, even though subjects moved faster in AR, we did not observe a significant difference for pointing performance between AR and VR headsets, which means that previously identified differences in depth perception between these platforms have no strong effect on interaction. Our results also help 3D user interface designers understand how changes in target depth affect users’ performance in different movement directions in AR and VR.
Our second paper is called “Effects of 3D Rotational Jitter and Selection Methods on 3D Pointing Tasks” Please find the abstract below:
3D pointing devices rely on 3D trackers and are thus subject to fluctuations
in the reported pose, i.e., jitter. In this work, we explored
how different levels of rotational jitter affect pointing performance
and if different selection methods can mitigate the effects of jitter.
Towards this, we designed a Fitts’ Law experiment with three selection methods. In the first method, subjects used a single controller to position and select the object. In the second method, subjects
used the controller in their dominant hand to point at objects and
the trigger button of a second controller, held in their non-dominant
hand, to select objects. Finally, subjects used the controller in their dominant hand to point the objects and pressed the space bar on a keyboard to select the object in the third condition. During the
pointing task, we added five different levels of jitter: no jitter, 0:5°,
1°, and 2° uniform noise, as well as White Gaussian noise with
1° standard deviation. Results showed that the Gaussian noise and
2° of jitters significantly reduced the throughput of the participants.
Moreover, subjects made fewer errors when they performed the experiment with two controllers. Our results inform the design of 3D
user interfaces, input devices and interaction techniques