Research

Smart homes, read from sensors

Both pieces of research are about recognising what someone is doing at home from motion and door sensors, with no cameras. Each has a page that explains it in plain language.

M.S. thesis, University of Georgia, 2026
Reading smart-home sensors four ways
Multimodal Self-Supervised Activity Recognition on CASAS Smart-Home Sensor Streams
A smart home can tell what a resident is doing without a camera, just from motion and door sensors clicking on and off. My thesis asks how best to read those sparse event logs, and finds that the answer depends on the house.
Complete. PDF to download.
Read the thesis explained
Conference paper
FACET
FACET: Factorized Ambient-Sensor Experts with Local Competence-Weighted Fusion for Smart-Home Activity Recognition
A paper on recognising everyday activities from ambient smart-home sensors, submitted to a conference and currently under review.
Submitted to a conference, under review
About the paper

Reading smart-home sensors four ways

FACET