Candidate: Pantea Habibi
Date and Time: Friday, November 1, 2024; 3:30 PM - 6:00 PM
Location: ERF 2068

Abstract:
When interacting with large displays, virtual reality, or augmented reality, people often use mid-air gestures. During these touchless interactions, they may need to alternate between hands or use both hands simultaneously. Most individuals have a hand preference, and preferred and non-preferred hands in lateralized individuals perform differently due to cerebral hemispheric specialization. For example, in right-handed individuals, the right hand is typically better at sequential motor tasks - those requiring (sensory) feedback control - whereas the left hand is better at parallel processing or open-loop behavior. In this thesis, we examine how lateral asymmetry impacts user performance in touchless input. My research provides an empirical investigation to characterize touchless performance when using preferred, non-preferred, and both hands. It explores binding the related tasks sequentially and in parallel, as well as different phrase structures, considering how much muscular tension is involved.

In this work, we compared preferred and non-preferred hand performance in simple and compound tasks, such as pointing and select-and-dragging, respectively. In right-handed individuals, we did not find any significant performance differences between their preferred and non-preferred hands in touchless pointing or dragging. However, between-hand performance differences were significant when using mouse and stylus input.

Additionally, we studied how phrasing with both hands can support volumetric multi-selection in touchless input. We present an empirical comparison of mid-air multi-selection techniques spanning dominant and non-dominant hand actions. The symmetric-synchronous and high-tension asymmetric-asynchronous bimanual techniques were not significantly different in terms of efficiency, and accuracy for within-reach targets. Furthermore, there were no significant differences between low and high-tension asymmetric-asynchronous bimanual techniques. The symmetric-synchronous bimanual technique was most preferred by users.

Lastly, we explored how different phrase structures in bimanual phrasing impact compound task performance, such as touchless multi-selection and manipulation (e.g., drag-and-drop). Our findings revealed that while accuracy remained consistent across different tension levels, higher tension led to faster completion times through continuous gesture engagement. Interestingly, participants preferred half-tension phrasing for its ergonomic balance, as it minimized fatigue without compromising efficiency. These empirical findings can inform the design of future bimanual and multimodal interaction techniques involving touchless input.