Research Projects

My research integrates computational linguistics and machine learning to build human-centered AI systems. These projects demonstrate the intersection of domain expertise in linguistics, cognitive science, and speech processing with modern ML/AI techniques.

Emotion-Aware Conversational AI for Dementia Care

Role: AI/ML Lead Project funded by UCOP Noyce Initiative

Overview

Leading the development of a clinically-validated AI pipeline for managing dementia-related agitation through therapeutic conversational interactions. This project creates the first AI system that combines real-time emotion recognition with expert-designed therapeutic response generation for the 55+ million people worldwide living with dementia.

Technical Contributions

Linguistics & Cognitive Science Integration

Technologies: Python, PyTorch, Transformers, OpenAI Whisper, wav2vec2, Azure Cognitive Services, Streamlit, HuggingFace, SentenceTransformers

Understanding Emotion in Discourse: From Recognition to Generation-Informed Insights

Role: Principal Investigator TMLR Manuscript in Preparation

Overview

Systematic analysis of emotion recognition in conversation (ERC), moving beyond "black box" accuracy to understand how models work. Achieved state-of-the-art text-only performance on IEMOCAP using strictly causal context, surpassing prior methods that exploit future utterances.

Technical Contributions

Linguistics & Cognitive Science Integration

Technologies: Python, PyTorch, CUDA, Transformers, RoBERTa, LSTM, Scikit-learn, HuggingFace Accelerate, SenticNet

Additional Projects

LLM Alignment for Empathetic Responses

Investigating Direct Preference Optimization (DPO) for training language models to generate empathetic responses in healthcare contexts. Developing annotation frameworks for preference data collection.

Technologies: Python, Transformers, DPO, Preference Learning

Multimodal Team Communication Analysis

Contributed to MultiCAT, a comprehensive annotation framework for multimodal team communication, published at NAACL 2025. Developed annotation schemas for verbal and non-verbal communication patterns in collaborative settings.

Technologies: Multimodal annotation, Inter-rater reliability analysis

Big Data Phonetics: Korean Stop Hyperarticulation

Applied automated acoustic analysis to 100,000+ tokens from Korean broadcast speech, demonstrating how speakers hyperarticulate phonetic cues in lexically confusable contexts. This work bridges corpus linguistics with speech technology.

Technologies: Python, Praat scripting, Forced alignment, Statistical modeling

Technical Skills

Programming Languages

Python, R, C++, Bash

ML/AI Frameworks

PyTorch, TensorFlow, HuggingFace Transformers, Scikit-learn

NLP & Speech

LLMs (GPT-4, Claude), Whisper, wav2vec2, BERT/RoBERTa, Praat

Computer Vision

MoveNet, OpenCV, Pose Estimation

Cloud & Hardware

NVIDIA GPUs (Academic Grant), Saturn Cloud, Azure Cognitive Services

Research Methods

Experimental design, Statistical analysis, IRR, Clinical validation