01
Wearable PPG
Ring- and wrist-form photoplethysmography for HR, HRV, RR, SpO₂, and BP under rest, daily-life motion, and running.
Researcher · Wearable physiological sensing
Yi Wang
王熠Wang Yi
I build wearable-compatible physiological AI — ring and wrist PPG, cross-modal cardiac waveform reconstruction, and non-EEG sleep staging — with a consistent focus on motion robustness and reliability-aware reporting for signals that have to survive daily life.
- 01
- published dataset
Nature Sci. Data - 03
- papers & thesis
2026 - 02
- interactive demos
live online
Research themes
02
PSG sleep staging
Cardiorespiratory and EEG/EOG-based sleep architecture modeling with compact, parameter-efficient backbones.
03
Cross-modal reconstruction
Mapping between cardiac waveforms (ECG ↔ PPG, BCG → ECG) to bridge sensor classes and exploit existing clinical datasets.
04
State-space backbones
Mamba / VMamba and other selective SSMs as a parameter-efficient methodological thread across multi-scale physiological signals.
Selected publications
01
2026
HealthRing: Physiology Dataset for Health Sensing on Rings
Nature Scientific Data · published
02
2026
StoryEcho: A Generative Child-as-Actor Storytelling System for Picky-Eating Intervention
ACM UIST 2026 · in submission · co-authored
03
2026
CITRUS: Continuous Insight into Trusted Rhythms of Universal Signals
MPhil thesis · HKUST(Guangzhou) · defending May 2026
Projects & demos
AI + interactive web demo
rppgbench.online
Browser-side interactive benchmark for remote-PPG models. Upload a face video, see POS / CHROM / DeepPhys / TSCAN / PhysFormer HR estimates side-by-side with stratified MAE.
Open project →
AI + interactive demo · UIST 2026
SGGG
Demo system behind StoryEcho, our UIST 2026 submission on a generative child-as-actor storytelling system for picky-eating intervention. Code and write-up at github.com/b55585wy/SGGG.
Open project →
MPhil thesis · CITRUSDefending May 2026
Continuous Insight into Trusted Rhythms of Universal Signals
My thesis threads three studies on continuous physiological sensing — ring PPG for cardiovascular vital signs across rest, daily activity, and running; an exploratory ECG ↔ PPG cross-modal waveform bridge; and a compact PSG-side sleep-staging baseline — connected by parameter-efficient state-space architectures.
- 54
- participants
- 49.76 h
- ring recordings
- 0.898
- error-risk AUC
Get in touch
I am applying for PhD positions for 2026/27 in wearable sensing, biomedical AI, and mobile health. If my work fits your group, I would love to hear from you.