Our group investigates problems at the intersection of machine learning, sensing systems, and human-centered AI. Our research spans multiple domains with the shared goal of building robust, privacy-preserving, and adaptive intelligent systems.
We apply symbolic modeling and generative AI to infer human behaviors in smart homes and assistive environments, focusing on activities of daily living, sleep, and mobility.
We develop robust and personalized learning algorithms that enable collaboration across decentralized data silos without sharing raw data. Applications include education analytics, health monitoring, and sensor networks.
We create contrastive and interpretable models for biological data (e.g., kinase-substrate prediction, trophic dynamics) by integrating domain knowledge and multi-source signals.
We build resilient MAC protocols and ML-based analytics to support sensing in underwater, remote, and low-power environments.