Collaborative Research: HCC: Medium: HCI in Motion -- Using EEG, Eye Tracking, and Body Sensing for Attention-Aware Mobile Mixed Reality

点击翻译按钮获取中文摘要

Mobile, wireless, headsets for virtual and augmented reality, such as the Meta Quest 2 and Microsoft HoloLens-2, are becoming more widely used in many applications beyond video games, such as training, construction, and medicine. However, wearing these head-worn goggles while walking can make some people feel sick or distracted, which has even led to injury in some cases. This effect is similar to texting while walking, but potentially worse because a person's entire periphery can be filled with distracting media elements. While previous research has investigated these issues when users are standing still or seated, it is unclear how problems unfold and how they can be prevented while users are in motion. Specifically, this project will investigate how and why virtual and augmented reality headsets affect attention and feelings of sickness. First, this work will record data, such as heart rate, brain waves, and the direction users are turning their eyes to, while they are wearing virtual and augmented reality headsets and walking. Secondly, this project will develop ways to reduce sickness and distraction while walking with virtual and augmented reality headsets. This work will improve the safety of mobile virtual and augmented reality headsets, products that virtually all big technology companies today heavily invest in as possible companions or replacements to smartphones. This project will be introduced in courses and research mentorship projects at The University of Texas at San Antonio and the University of California at Santa Barbara, to advance research training of both undergraduate and graduate students. Considering that both universities and research teams have a history of supporting many underrepresented minority students, it is expected that the educational value of this project will be high, especially in terms of recruiting and mentoring women and underrepresented minority students.There is an increasing prevalence of mobile, immersive interfaces (e.g., mobile Virtual Reality(VR) / Augmented Reality (AR)) that may affect users' cognitive capacities and situational awareness, potentially leading to physical harm (e.g., impaired task performance, tripping over physical obstacles in VR, unsafe street crossings while seeing advertisements in AR). The landscape of human-computer interaction has expanded from fairly well standardized stationary office configurations to more varied mobile and immersive settings involving active body movements (mobile and situated computing, AR, mobile VR) and simulated first-person perspective changes and motion experiences (immersive computing). To make matters worse, compared to more standardized platforms such as desktop and laptop UIs, tablet and smartphone interfaces, individual differences among users have a much bigger usability impact in context-driven surround-focus usage scenarios found in mobile AR/VR. For example, motion sickness (i.e., cybersickness) in VR is known to inflict symptoms of widely varying severity, depending on the individual user. One serious consequence is that interaction designers have difficulties providing engaging general experiences that are universally usable by a wide variety of users. Despite the increasing prevalence of immersive technologies and their pitfalls, the precise cognitive and physiological mechanisms at play when 'computing in motion' are not well understood. This work is aimed at filling this knowledge gap. The specific objectives are: 1) to assess the cognitive effects of interacting with mobile AR/VR while users are walking, 2) to provide automated tools to effectively reduce the cognitive demand of mobile AR/VR, and 3) to make mobile AR/VR safer and more usable. Based on preliminary data, the central hypothesis is that through multi-modal sensing combined with machine learning approaches, mobile AR/VR applications can learn the characteristics of user behavior and provide real time adaptations that will reduce user error, increase ease of use, improve task performance, and reduce the impact of physical hazards. This work will improve the safety of mobile AR and mobile VR, paradigms that virtually all big technology companies today heavily invest in as a possible follow-up paradigm to the smartphone platform. Educational impact will occur through incorporation of research outcomes into undergraduate and graduate courses offered at The University of Texas at San Antonio and the University of California at Santa Barbara, and research training and mentorship opportunities for both undergraduate and graduate students. The courses include Machine Learning, Deep Learning for Visual Computing, Human-Computer Interaction, and Mobile Application Programming. Because our project integrates a topic of high social impact with cutting edge machine learning and human-computer interaction research along with proven successful mentorship strategies, the educational impacts of the project will be high, especially in terms of recruiting and mentoring women and underrepresented minority students.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.