Attune
UIST 2026
Understanding operator attention during multi-robot supervision.
Puqi Zhou
- Human–Robot Interaction
- Multi-Robot Systems
- Human Attention
- Shared Autonomy
I'm a Computer Science Ph.D. student at George Mason University, advised by David Porfirio.
My research focuses on human-centered multi-robot systems: I model human attention and behavior, and use these models to design adaptive interfaces and shared autonomy for real-world robot teams.
Seeking research internships · Spring / Summer 2027
Research Directions
-
Understand Human attention in multi-robot supervision
Characterize when, where, and why operators shift attention across robots.
-
Model Human attention prediction
Model how robot motion, visual information, and task context shape attention over time.
Predicting Attention in Multi-Robot Supervision
Current workModeling how robot actions and visual context predict where operators attend.
Earlier work
Theia
MobiSys 2024
Gaze and perception in mixed-reality video delivery.
-
Design Adaptive interfaces and shared autonomy
Use models of human state and task context to decide what robots should communicate or handle autonomously.
Shared Social Autonomy for Robot Teleoperation
Current workDesigning autonomy that handles low-level social behavior while preserving operator control.
Earlier work
Collaborative learning in VR
CHI 2023
Shared video control for collaborative learning.
Volumivive
IEEE VRW 2023
Authoring interactive volumetric video.
-
Deploy Human–robot teams in real-world workflows
Study how multi-robot systems can fit into professional workflows such as public safety and search.
Multi-robot video sensemaking
CHI 2026
Video interfaces designed with public-safety professionals.
Ground Robots for Urban Missing-Person Search
Current workStudying how ground robots can complement professional search teams in urban environments.
News
All newsAttune accepted to UIST 2026.
Received the 2026–2027 Doctoral Research Scholarship at George Mason University.
3rd place, Mason AI Day Graduate Poster Competition.
1st place presenter, PhD Symposium.
Our work on multi-robot video sensemaking accepted to CHI 2026.
Selected Publications
All publications
UIST 2026
Attune: A Self-Annotation Tool for Understanding Robot Operator Attention Profiles
Attune lets robot operators replay their gaze and annotate why their attention shifted. We combine robot-camera video, eye tracking, and AI-assisted annotation to build interpretable attention profiles for multi-robot supervision.
CHI 2026
Designing Multi-Robot Ground Video Sensemaking with Public Safety Professionals
Working with public-safety professionals, we studied how to make video from ground robots useful in their workflows. The resulting interface brings together multiple video streams, event explanations, and tools for reviewing events.