I am a PhD candidate at the
Pervasive HCI Lab,
Department of Computer Science and Technology,
Tsinghua University,
advised by Professors Chun Yu and Yuanchun Shi.
I received my bachelor’s degree in computer science from Tsinghua University.
I build Cognitive Mirrors:
my work develops contextually intelligent AI systems that mirror human contexts, goals, and reasoning processes through shared representations, enabling people and AI to reflect, align, and move forward together in complex tasks.
Publications
OpenCD: Empowering Diagnosis of Children's Mathematical Cognition through Open-ended Multimodal Tasks
Zhi Zheng, Chun Yu, Weihao Chen, Minzheng Song, Binglin Liu, Jianyang Liu, Shiyi Wang, Xutong Wang, Jie Cai, Yuanchun Shi
CHI '26
OpenCD uses MLLMs to diagnose children's mathematical cognition from open-ended multimodal tasks, such as how a child draws and explains "46", turning rich, messy learning behaviors into interpretable cognitive evidence.
EchoMind: Supporting Real-time Complex Problem Discussions through Human-AI Collaborative Facilitation
Weihao Chen, Chun Yu, Yukun Wang, Meizhu Chen, Yipeng Xu, Yuanchun Shi
CSCW '25
EchoMind visualizes live conversations as real-time issue maps, creating a shared attention space where people and AI can track ideas, align focus, and advance complex discussions together.
RingMind: Interpreting Multimodal Behavior from Ring-Based Speech and Gesture using Large Language Models
Weihao Chen, Yukun Wang, Meizhu Chen, Zhe He
UbiComp '25 Challenge
RingMind leverages speech and co-speech gestures sensed from a smart ring, turning rich human signals into multimodal context for LLM-based understanding.
CATIA was deployed on 36 users' own smartphones in a 7-day in-the-wild study, using everyday context to collaborate on text entry with 91.58% goal inference accuracy, 82.36% no-edit acceptance, and two observed mental models.
From Gap to Synergy: Enhancing Contextual Understanding through Human-Machine Collaboration in Personalized Systems
Weihao Chen, Chun Yu, Huadong Wang, Zheng Wang, Lichen Yang, Yukun Wang, Weinan Shi, Yuanchun Shi
UIST '23
LangAware enables in-situ collaboration around context: people and machines co-construct shared contextual concepts, improving rule-creation success to 87.50% while reducing the need for language repairs.
BiTipText: Bimanual Eyes-Free Text Entry on a Fingertip Keyboard
Zheer Xu, Weihao Chen, Dongyang Zhao, Jiehui Luo, Te-Yen Wu, Jun Gong, Sicheng Yin, Jialun Zhai, Xing-Dong Yang
CHI '20
BiTipText makes text entry always available through bimanual, eyes-free fingertip keyboards, using behavior modeling to optimize how people type on the hand itself.
A toolkit built around Handy Prompt, a human-friendly .hprompt format that makes LLM workflows readable, reusable, and runnable across Python, CLI, and editors.
A typed Socket.IO library that brings Pydantic validation to real-time events. It helps replace fragile ad-hoc message schemas with explicit, typed event data.
A lightweight daemon manager for running ordinary commands as background services without Docker or heavyweight infrastructure, with logging and log rotation built in.
A toolkit for matrix-sensor experiments, supporting real-time streaming, offline processing, and visualization for rapid sensing and interaction prototyping.
Pervasive HCI Lab,
Department of Computer Science and Technology,
Tsinghua University,
advised by Professors Chun Yu and Yuanchun Shi.
I received my bachelor’s degree in computer science from Tsinghua University.
I build Cognitive Mirrors:
my work develops contextually intelligent AI systems that mirror human contexts, goals, and reasoning processes through shared representations, enabling people and AI to reflect, align, and move forward together in complex tasks.