Erdem Murat

• Advisor: Prof. Lap-Fai (Craig) Yu

Conducted research in virtual reality, gameplay design/engineering, machine learning, and systems analysis with the aim of publishing in the top academic conferences.

• As a member of the lab, I have been able to:

• Receive mentorship from Prof. Yu.
• Collaborate with lab members.
• Access equipment and software required in research & development.
• Attend seminars, conferences, and  related events.

• Created virtual reality hubs for the  Global Co Lab to host conferences and present fieldwork addressing social issues.

• Optimized the prior build by 70% by utilizing functionalities in Blender, Spoke and Photoshop to reduce rendering workload.
• Instructed and mentored 3 teams towards developing educational and entertaining VR environments for the UN Sustainable Development Goals (SDG) Metaverse Competition

• Presented work at the United Nations Science Technology and Innovation Forum.

• Developed a comprehensive, 5-day curriculum with 30 hours of educational content to teach virtual reality development concepts to 20 students.
• Harnessed knowledge based on emerging VR technology through attending lectures, conducting research, and attending seminars & conferences to construct content from the field of VR.
• Integrated advice from the Mozilla Hubs team to further improve the quality and comprehensiveness of teaching materials.

Erdem Murat, Liuchuan Yu, Siraj Sabaj, Haikun Huang, Lap-Fai Yu

ACM UIST 23’ Under review
• Proposed a novel application that uses machine learning to predict users’ difficulty perception in virtual reality games.
• Addressed issues in VR game design by adapting decades of non-VR game design and gameplay engineering research.
• Obtained IRB certificate, collected data by conducting 70+ user studies, and trained a recurrent neural network to understand recurring relationships between game parameters, user data, and gameplay data.

Ruizhi Cheng, Erdem Murat, Lap-Fai Yu, Songqing Chen, Bo Han

Under Review

• Proposed a novel analytic method that combines qualitative and quantitative analysis with end-to-end network measurements to understand user experience in VR education and detect bottlenecks in system performance.
• Deployed a Mozilla Hubs server-client system with custom JavaScript scripts to record client-side performance through an API.
• Used Python and Jupyter Notebook to compute, visualize and analyze systems data.
• Gave a VR lecture to 23 Graduate students and later assisted a Meta researcher in using my platform for a seminar.

View Paper Here

• Developed a system to address a limitation in VR development research and proposed a solution that improves user experience through machine learning to automate the manipulation of experience-affecting parameters in VR.
• Deployed a testing system to record the effect of parameter manipulation on user experience based on a 10-point rating scale of difficulty, fun, and stress.
• Used collected data to train a machine learning regressor the relations between game parameters and perceived difficulty with low prediction error (26%).
• Obtained IRB certificate for social and behavioral research and conducted user studies with over 50 participants