About me
I am a Ph.D. candidate in Computer Science at the University of Texas at Dallas, advised by Prof. Yapeng Tian. My research is on AI for situated interaction: multimodal systems that understand ongoing human activity from a first-person point of view, with the long-term goal of proactive, personalized spatial AI assistants. I build causal egocentric perception models, benchmarks that test whether models interpret people, objects, and tasks in context, and methods that teach multimodal models which evidence to trust. With collaborators in HCI, accessibility, and developmental science, I apply this work to assistive systems for people with low vision and to gaze analysis for young autistic children.
I teach my own section of ECS 1100 at UT Dallas as instructor of record, completed the Graduate Teaching Certificate, and have mentored five undergraduate researchers. Prior to UTD, I received a Bachelor of Science in Mathematics from Sichuan University.
I am on the 2026–27 faculty job market.
News
Selected Publications
Building multimodal AI that understands ongoing human activity and knows when its help would be useful.
Perceiving What Matters
Causal egocentric perception that grounds first-person video in user attention: where the user is looking and which objects deserve to be surfaced.
Interpreting in Context
Models and benchmarks that interpret people, objects, and tasks in their social, physical, and procedural context.
Knowing Which Evidence to Trust
Evaluation and training methods that expose and reduce failures when inputs are uninformative or conditions depart from training.
Projects
