Experience

Research Projects

Measuring Learning Engagement in Multi-Person Virtual Reality

Human Factors & Simulation Lab, University of Oklahoma · Aug 2024 – Present
NSF CAREER project: “Non-text-based smart learning in fully immersive multi-person virtual reality using multimodal analysis of physiological measures.” PI: Dr. Ziho Kang

In this project, people in different places join one virtual classroom. We use eye movements, hand interactions, and brain activity to measure and predict how engaged they are in learning.

Helped design the multi-person VR experiments and build the task scenes in Vizard and Godot. Ran the experiments with more than 60 participants and recorded fNIRS, eye-tracking, and behavior data.

Single-person VR semantic network task — Left: the single-person VR semantic network task. Right: the multi-person version. Each colored laser shows where a participant is looking in real time.

Multi-person VR semantic network task — Left: the single-person VR semantic network task. Right: the multi-person version. Each colored laser shows where a participant is looking in real time.

Led the fNIRS data analysis, from preprocessing and filtering to motion artifact correction.
Found that the eye-tracking cameras in VR headsets add periodic infrared noise to fNIRS signals. Developed a filtering method based on FFT and seasonal-trend decomposition (STL) to remove it. The method needs no hardware change and works on data that has already been collected.

fNIRS signal before filtering — fNIRS signal from one optode before (left) and after (right) filtering. The regular peaks on the left come from the eye-tracking cameras.

fNIRS signal after filtering — fNIRS signal from one optode before (left) and after (right) filtering. The regular peaks on the left come from the eye-tracking cameras.

Built time-window measures from head, hand, and gaze data. These show how behavior changes during a task.
Outcome: one first-author IEEE conference paper (read the paper) and one accepted HFES paper. A journal paper on the filtering method is in preparation.

Human Perception of Adaptive Cruise Control (ACC)

Human Factors & Simulation Lab, University of Oklahoma · Mar 2024 – 2026 · Project Lead

Adapted a validated ACC simulation model in MATLAB. Compared an electric vehicle and a gas vehicle, each following a gas vehicle, in two scenarios: a highway merge with hard braking, and stop-and-go traffic.
Introduced three measures of driver perception: perceived safety, consistency, and comfort. The electric vehicle scored higher on all three in both scenarios.
Outcome: first-author paper in Traffic Injury Prevention (2026). Read the paper

ACC car-following system — Overview of the ACC system. The following car uses its front sensor to track the car ahead and adjust its speed.

University of Virginia · Mar 2023 – May 2023

Helped build a probabilistic model of login failures on the eBenefits / VA.gov website.
Implemented and checked the model algorithms.
Showed that the DS Logon path improves the login success rate.
Outcome: co-authored paper at the AHFE 2023 International Conference.

Login model for the eBenefits website — Model of the login states on the eBenefits website.

Work Experience

Research Assistant

Human Factors & Simulation Lab, University of Oklahoma · Mar 2024 – Present

Work on VR experiments and driving simulation studies.
Process and analyze fNIRS, eye-tracking, and behavior data in Python, R, and MATLAB.
Write conference and journal papers with the research team.

Teaching Assistant

School of Industrial and Systems Engineering, University of Oklahoma · Aug 2024 – Present

Teach lab sessions, grade coursework, and hold office hours. See Teaching.

Technical Support Engineer Intern

iFLYTEK Co., Ltd., Hefei, China · Jul 2021 – Sep 2021

Analyzed behavior and feedback data from over 1,000 users of online education platforms.
Found common usability problems and shared them with the design team.
Maintained user databases with Navicat and Fiddler.

Skills

Programming: Python, R, MATLAB
Simulation: Arena, PTV Vistro
VR development: Vizard, Godot
Design and analysis tools: Figma, SAS, Origin
Data types: fNIRS, eye tracking, VR motion data

Jingan Peng