A visual analytics study based on 3D surface plots that jointly encode sugar intake dose, temporal evolution, and glucose response magnitude to characterize free-living postprandial interstitial glucose responses.

Dual-encoder demonstration Retrievers Architecture (DoRA) is a model to retrieve in-context demonstrations for out-of-domain (OOD) tasks.

This repository contains the source code for a novel data pipeline designed to extract time-series laboratory results and unstructured clinical notes corresponding to specific hospital admissions, along with comprehensive evaluations using large language models for multi-morbidity disease phenotype classification.

ExDoRA is a novel LLM-based framework to enhance the interpretability of LLMs and improve out-of-domain transferability for unseen tasks using computational intelligence techniques.

Ubicomp lab based in Kyoto, Japan

GEM is a Graph Attention Encoder for Multi-task Depression Severity Detection in Multi-party Conversations.

A watchface app that uses werewolf to nudge people for more steps

A novel framework which is specifically designed to use a PLM's knowledge in structure and semantic modelling in multi-party conversations to perform depression detection.