Data Science & Machine Learning researcher
PhD in Computer Science
A resource-aware multimodal sensing system for on-device eating detection and caloric estimation.
A fully on-device smart cane for proactive floor-level trip-hazard detection with gyroscope-gated vision-language inference.
Transformer-based eating detection and caloric modeling with privacy-preserving visual obfuscation.
Privacy-preserving cartoonization for wearable camera–based eating context recognition.
Wearable thermal-RGB sensing for real-time smoking and eating detection with smartwatch confirm-refute EMA validation.
BMI-inclusive energy expenditure estimation on wrist-worn sensors.
Longitudinal ML pipeline reveals overeating patterns.
Knowledge-graph framework for lifestyle pattern mining.
Dual-lead ECG-PPG patch for continuous stress sensing.
Low-latency, energy-efficient intake detection on wrist IMU.
Opportunities and challenges of wearables for eating research.
Explainable ML for next-day stress prediction in pregnancy.
RadViz-t implementation for large-scale data visualization.
Deep neural networks for activity recognition.