The primary objective of this research is to understand how users manipulate virtual objects in augmented reality using multimodal interaction (gesture and speech) and unimodal interaction (gesture). Through this understanding, natural-feeling interactions can be designed for this technology. These findings are derived from an elicitation study employing Wizard of Oz design aimed at developing user-defined multimodal interaction sets for building tasks in 3D environments using optical see-through augmented reality headsets. The modalities tested are gesture and speech combined, gesture only, and speech only. The study was conducted with 24 participants. The canonical referents for translation, rotation, and scale were used along with some abstract referents (create, destroy, and select). A consensus set of gestures for interactions is provided. Findings include the types of gestures performed, the timing between co-occurring gestures and speech (130 milliseconds), perceived workload by modality (using NASA TLX), and design guidelines arising from this study. Multimodal interaction, in particular gesture and speech interactions for augmented reality headsets, are essential as this technology becomes the future of interactive computing. It is possible that in the near future, augmented reality glasses will become pervasive.

This research establishes a better understanding of the syntax choices in speech interactions and of how speech, gesture, and multimodal gesture and speech interactions are produced by users in unconstrained object manipulation environments using augmented reality. The work presents a multimodal elicitation study conducted with 24 participants. The canonical referents for translation, rotation, and scale were used along with some abstract referents (create, destroy, and select). In this study time windows for gesture and speech multimodal interactions are developed using the start and stop times of gestures and speech as well as the stoke times for gestures. While gestures commonly precede speech by 81 ms we find that the stroke of the gesture is commonly within 10 ms of the start of speech. Indicating that the information content of a gesture and its co-occurring speech are well aligned to each other. Lastly, the trends across the most common proposals for each modality are examined. Showing that the disagreement between proposals is often caused by a variation of hand posture or syntax. Allowing us to present aliasing recommendations to increase the percentage of users' natural interactions captured by future multimodal interactive systems.

The role of affect has been long studied in human–computer interactions. Unlike previous studies that focused on seven basic emotions, an avatar named Diana was introduced who expresses a higher level of emotional intelligence. To adapt to the users various affects during interaction, Diana simulates emotions with dynamic facial expressions. When two people collaborated to build blocks, their affects were recognized and labeled using the Affdex SDK and a descriptive analysis was provided. When participants turned to collaborate with Diana, their subjective responses were collected and the length of completion was recorded. Three modes of Diana were involved: a flat-faced Diana, a Diana that used mimicry facial expressions, and a Diana that used emotionally responsive facial expressions. Twenty-one responses were collected through a five-point Likert scale questionnaire and the NASA TLX. Results from questionnaires were not statistically different. However, the emotionally responsive Diana obtained more positive responses, and people spent the longest time with the mimicry Diana. In post-study comments, most participants perceived facial expressions on Diana’s face as natural, four mentioned uncomfortable feelings caused by the Uncanny Valley effect.

Visual notifications are omnipresent in applications ranging from smart phones to Virtual Reality (VR) and Augmented Reality (AR) systems. They are especially useful in applications where users performing a primary task have to be interrupted to react to external events. However, these notifications can cause disruptive effects on the performance of users concerning their currently executed primary task. Also, different notification placements have been shown to have an influence on response times, as well as e.g. on user perceived intrusiveness and disruptiveness. We investigated the effects and impacts of four visual notification types in AR environments when the main task was performed (1) in AR and (2) the real world. We used subtitle, heads-up, world space, and user wrist as notification types. In a user study, we interrupted the execution of the main task with one of the AR notification types. When noticing a notification, users responded to it by completing a secondary task. We used a Memory card game as the main task and the pressing of a correctly colored button as the secondary task. Our findings suggest that notifications at a user’s wrist are most suitable when other AR elements are present. Notifications displayed in the World are quick to notice and understand if the view direction of a user is known. Heads-up notifications in the corner of the field-of-view, as they are primarily used in smart glasses, performed significantly worse, especially compared to Subtitle placement. Hence, we recommend to use different notification types depending on the overall structure of an AR system.

As augmented reality (AR) technology and hardware become more mature and affordable, researchers have been exploring more intuitive and discoverable interaction techniques for immersive environments. This paper investigates multimodal interaction for 3D object manipulation in a multi-object AR environment. To identify the user-defined gestures, we conducted an elicitation study involving 24 participants and 22 referents using an augmented reality headset. It yielded 528 proposals and generated a winning gesture set with 25 gestures after binning and ranking all gesture proposals. We found that for the same task, the same gesture was preferred for both one and two-object manipulation, although both hands were used in the two-object scenario. We present the gestures and speech results, and the differences compared to similar studies in a single object AR environment. The study also explored the association between speech expressions and gesture stroke during object manipulation, which could improve the recognizer efficiency in augmented reality headsets.

While the visual search for targets in a complex scene might benefit from using augmented-reality (AR) head-mounted display (HMD) technologies by helping to efficiently direct human attention, imperfectly reliable automation support could manifest in occasional errors. The current study examined the effectiveness of different HMD cues that might support visual search performance and their respective consequences following automation errors. A total of 56 participants searched a three-dimensional environment containing 48 objects in a room, in order to locate a target object that was viewed prior to each trial. They searched either unaided or assisted by one of the three HMD types of cues: an arrow pointing to the target, a plan-view minimap highlighting the target, and a constantly visible icon depicting the appearance of the target object. The cue was incorrect in 17% of the trials for one group of participants and 100% correct for the second group. Through both analysis and modeling of both search speed and accuracy, the results indicated that the arrow and minimap cues depicting location information were more effective than the icon cue depicting visual appearance, both overall, and when the cue was correct. However, there was a tradeoff on the infrequent occasions when the cue erred. The most effective AR-based cue led to a greater automation bias in which the cue was more often blindly followed without careful examination of the raw images. The results speak to the benefits of AR and the need to examine potential costs when AR-conveyed information may be incorrect because of imperfectly reliable systems.