PhD Oral Exam - Yashas Joshi, Computer Science

When studying for a doctoral degree (PhD), candidates submit a thesis that provides a critical review of the current state of knowledge of the thesis subject as well as the student’s own contributions to the subject. The distinguishing criterion of doctoral graduate research is a significant and original contribution to knowledge.

Once accepted, the candidate presents the thesis orally. This oral exam is open to the public.

Abstract

Among many virtual reality interactions, the locomotion dilemma remains a significant impediment to achieving an ideal immersive experience. The physical limitations of tracked space make it impossible to naturally explore theoretically boundless virtual environments with a one-to-one mapping. Synthetic techniques like teleportation and flying often induce simulator sickness and break the sense of presence. Therefore, natural walking is the most favored form of locomotion. Redirected walking offers a more natural and intuitive way for users to navigate vast virtual spaces efficiently. However, existing techniques either lead to simulator sickness due to visual and vestibular mismatch or detract users from the immersive experience that virtual reality aims to provide.

This research presents innovative techniques and applications to enhance the user experience by expanding walkable, physical space in Virtual Reality. The thesis includes three main contributions. The first contribution proposes a mobile application that uses markerless Augmented Reality to allow users to explore a life-sized virtual library through a divide-and-rule approach. The second contribution presents a subtle redirected walking technique based on inattentional blindness, using dynamic foveated rendering and natural visual suppressions like blinks and saccades. Finally, the third contribution introduces a novel redirected walking solution that leverages a deep neural network, to predict saccades in real-time and eliminate the hardware requirements for eye-tracking.

Overall, this thesis offers valuable contributions to human-computer interaction, investigating novel approaches to solving the locomotion dilemma. The proposed solutions were evaluated through extensive user studies, demonstrating their effectiveness and applicability in real-world scenarios like training simulations and entertainment.