Soheil Sepahyar, Ph.D.

VR AR and AI researcher and lecturer at UMass Boston. Distance perception, foveated rendering, HCI, and robot learning with Isaac Sim.

Portrait of Soheil

About

I am a computer scientist working at the intersection of VR, AR, AI, graphics, and HCI. I build gaze-aware systems and study distance perception in immersive scenes. I also design human-in-the-loop data collection and evaluation for robot learning.

Stack highlights include C++, Python, C#, Unity, OpenGL, GLSL, and Maya. Current work focuses on eye-tracked foveated rendering and VR-to-Isaac Sim data pipelines.

Lecturer, UMass Boston
VR and AR
AI for XR and Robotics
HCI
Distance perception
C++ Python C#
VR and AR
Distance perception
C++ Python C#

Experience

Lecturer in Computer Science

University of Massachusetts Boston | Jan 2024 to Present
  • Courses include CS420, CS110, CS410, CS450, CS220, CS210, CS240
  • Developed materials and guided projects across theory and systems

PhD VR Researcher

Michigan Technological University | Jan 2019 to Dec 2023
  • Unity and C# for VR experiments and tools
  • Python for data analysis and experiment pipelines
  • OpenGL and GLSL for C and C++ graphics
  • User studies on distance perception in VR

Visteon Internships

Summers 2020, 2021, 2022
  • Product design lead intern for testing pipelines and data analysis
  • AI and ML for driver monitoring and dynamics
  • ADAS software with CNN depth perception

Education

PhD in Computer Science

Michigan Technological University | GPA 3.71

MS in Computer Science

Michigan Technological University | GPA 3.70

BS in Computer Software Engineering

Azad University, Central Tehran | GPA 3.74

Skills

  • Languages C++, Python, C#
  • Tech OpenGL, GLSL, Unity, Maya
  • Platforms Linux, Windows, macOS

Current Projects

Foveated Rendering and Distance Perception in VR

Current

I study how gaze-contingent foveated rendering changes egocentric distance judgments in VR using Meta Quest Pro eye tracking in Unity.

VR Data for Isaac Sim Robot Training

Current

I build a VR teleoperation pipeline to collect demonstrations in NVIDIA Isaac Sim and test if VR data improves robot learning on pick and place tasks.

Publications

VR Distance Judgments SAP 2022
ACM DL
Brightness on Distance Judgments SAP 2020
ACM DL
Sorting Algorithms Comparison ACAI 2019
ACM DL